AudiEvolve Documentation 2013

EV
Transforming the journey into the destination
ME310 2013 Final Documentation

Audi Team
Audi Corporation
Mechanical Engineering 310
Fall Design Documentation
Version: 10.12.2012.
Dept. of Mechanical Engineering
Stanford University
Stanford, CA 94305-4021
In Partnership With:
Aalto University, Helsinki, Finland
Final Documentation
Final Documentation
1. Prelude
1.1 Executive summary

With the advent of new robotic and
autonomous technologies, it was only
a matter of time until it crossed over to
vehicle design. The development in the
area of autonomous cars has already
reached levels that have made it possible for
companies like Google to manufacture and
test out self-driving cars, which have already
clocked over 200,000 miles in the US [10]. It
is not just the technology that is developing;
the entire support system is changing as
autonomous cars have been legalized in
three states in the US. Personal mobility
currently is often perceived as driving in a
vehicle. It is this perception that is about
to change; mobility will not just be about
driving anymore. Driving is not the primary
task that would be performed in autonomous
cars. These cars would offer a wide range of
options for various activities the user can do
inside the car while it is driving them around.
Team Audi Evolve consists of a combined
team of mechanical and electrical engineers,
product designers, industrial designers and
business school students from Stanford
University and Aalto University in Finland.
The team envisions a future where cabin
space designs of autonomous cars will have
transformed the journey into the destination.
The aim is to offer a solution for 2035 where
drivers can easily work and enjoy their free
time, while maintaining the pleasure of
driving. This makes it important to focus on
the issue of transitioning between various
activities and driving.
The important areas that have been identified
in the design space are psychological,
Final Documentation
Your autonomous car

is your personal driver
in the future!
physical and experiential. The team went
through several stages of prototyping to
explore the design space. The initial focus
was on a safe transition from autonomous to
manual driving mode. The main idea tested
was that of the steps involved in a safe
transfer of control to the user. The turning
point for this project came with a prototype
exploring a reconfigurable cabin space
design in cars. It was after this prototype that
the focus shifted on the actual experience
of the transition. Users who will adopt this
technology in the future are people who
will still want to drive along with performing
other activities in the cabin space during
autonomous mode. They will multi-task a
lot and always stay connected, therefore
changing positions frequently.
Since people will often be performing many
different activities while in the car, creating
an excellent riding experience is just as
important as an excellent driving experience.
The team is designing a system that will lead
to smooth and comfortable transitions not
only to and from driving mode, but also for
transitions between other activities. This
will also help people regain time lost in
commuting by creating a cabin space that
is adaptable to many activities and removes
tedious manual adjustments.
The designed system will give the user
complete flexibility to perform different
Final Documentation
activities in the cabin space. The car seat
interactions with the system Mode Initiators
and the steering mechanism are important
(MI) are intentional actions within reach like
components in the cabin space, which will
pulling the steering wheel to go into manual
affect the freedom of motion and ability to
driving mode and pushing on the steering
perform various activities in a comfortable
wheel to make it retract. The user can also
way. The team envisions that the system will
use Intentional Body Commands (IBC) like
consist of a smart chair and a retractable
leaning back to tilt the seat. The chair can
steering wheel. The main motivating factors
also give Adaptive Chair Reactions (ACR) to
for the retractable steering design were
movements of the users like rotating slightly
increasing mobility within the cabin space
when they reach for something at the back
and relieving the users of the responsibility
of the car.
of monitoring autonomous driving when in
This new cabin space experience allows for
autonomous mode. While the steering wheel
the driver to rotate 180 degrees to be able
is retracted and locked for autonomous
to socialize with other passengers. Having
mode, the tablet is enabled so that the user
only one seat in the front also increases
can interact with the windshield. When the
the mobility within the cabin, while also
steering wheel is not retracted it is free to
making the drivers seat very desirable.
rotate so that the driver can have manual
This configuration allows the driver to
control of the car. There are three different
better share the driving experience with
Figure 1.1.1: Final Prototype
Final Documentation
the passengers since it increases their view
and artificial intelligence. With so much
of the road as well. Our final prototype will
technological development in these major
give the experience of facilitated activities
fields there was bound to be a point when
through increasing mobility and effortlessly
these ideas crossover and start influencing
transitioning between activities for the driver.
automotive design. As technology develops
Going forward, the team will incorporate
further, it will lead to a complete paradigm
many of the findings uncovered in all of the
shift in the way cars and automotive
prototypes, research, and needfinding into
technology in general is perceived amongst
a functioning system. The development
the general population. Concepts which
described herein leads to a vision combining
were just science fiction a few years back,
many of the concepts explored during the
have started becoming real and achievable
last few months. Through a continued
in the near future. The development in
process of rapid prototyping and additional
the area of autonomous cars has already
user testing, the team plans to present a
reached levels which has made it possible
highly refined and fully functioning system
for companies like Google to manufacture
in June 2013.
and test out self driving cars which have
1.2 Project Background

Over the last 100 years personal vehicles
have become one of the most heavily
used means of transportation in developed
countries. In many ways, cars define who
we are, what we do and how we travel.
Along with increased mobility, cars have
brought about tremendous changes to the
way we live and interact. Over the period of
time, in addition to being just a medium of
transportation, cars have gone to become
a medium to experience the sheer pleasure
of racing around and driving in a machine
that is so responsive to the users actions.
Indeed a major component of the pleasure
of driving lies in the ability to drive on the
edge of control and experience the thrill of
being out there and taking the risk.
However, this entire landscape is changing
as we head towards the age of robots
already clocked over 200,000 miles in

the US. It is not just the technology that is
developing, but the entire support system
is changing with autonomous cars being
legalised in three states in the US.
Personal mobility currently is often perceived
as driving in a vehicle. It is this perception
that is about to change, mobility will not just
be about driving anymore. Driving is not the
primary task that would be performed in
autonomous cars. They would offer a wide
range of options with users being able to
perform various other activities inside the
car while it is driving them around.
The major hurdles that lie in the development
and adoption of this technology widely
are the costs associated with required
infrastructure development, legal issues
like liability in case of accidents and user
acceptability in general. The question that
Final Documentation
comes up is what would happen when
Since people will often be performing
this would actually launch in the market?
different activities while in the car,
Would people be willing to take the leap
creating an excellent riding experience is
of faith and trust their life on a computer
just as important as an excellent driving
controlled car? What kind of assurance
experience. Switching between manual
and trust development would be required
mode, autonomous mode and different
for widespread acceptance of autonomous
activities within autonomous mode, always
vehicles? These and a lot of other questions
includes a transition. It is these transitions
and issues are the prime concern of car
that the team hopes to address, as well as
manufacturers who are willing to step in
ensuring that the cabin has enough mobility
this field of autonomous car development.
in order to perform various activities. The

vision of Team Audi Evolve is to facilitate
Like any other new technology, it will take
activities for the driver through effortless
time to adopt and build trust on this system
transitions and increased mobility.
but it will happen eventually and once that

happens the possibilities that it opens up are
enormous. It is not presumptuous to assume
that by the year 2035, autonomous cars will
have become a common sight. Although
they offer relaxation and safety in driving,
it is still problematic for users to completely
rely on their autonomous functions. Car
designers and manufacturers then need to
focus on taking advantage of this whole new
opportunity to redefine urban mobility and
experience with cars.
The car will no longer be a medium for the
journey but it will be a productive location
where users can perform various activities
while being driven around autonomously.
1.3 Project Vision

Team Audi Evolve design team envisions a
future where people will still want to drive even
though driving will be only a secondary activity.
Autonomous driving allows spare time for the
driver and ensures increased productivity.
Final Documentation
1.4 Table of Contents

1. Prelude ............................................................................................. 3
1.1 Executive Summary ......................................................... 4
1.2 Project Background .......................................................... 6
1.3 Project Vision ................................................................... 7
1.4 Table of Contents ............................................................. 8
1.5 List of Figures .................................................................. 14
1.6 List of Tables .................................................................... 17
1.7 Glossary ........................................................................... 17
2. Context ........................................................................................... 25
2.1 Need Statement ................................................................ 26
2.2 Problem Statement .......................................................... 26
2.3 The Design Team
............................................................ 26
2.3.1 The Stanford Team ..........................................
28
2.3.2 The Aalto Team ................................................ 30

2.4 Corporate Sponsor ........................................................... 32
2.4.1 Corporate Liaison ............................................. 33
2.5 Teaching Team ................................................................. 33
2.6 Special Thanks ................................................................ 33
3. Design Requirements ................................................................... 35
3.1. Given Requirements ....................................................... 36
3.2 Functional Requirements ................................................
37
3.2.1 Functional Opportunities .................................
42
3.3 Physical Requirements ...................................................
43
3.3.1 Physical Opportunities ....................................
45
3.4 Business Opportunity
.....................................................
47
4. Design Development .................................................................... 49

4.1 Future Assumptions
........................................................ 50
4.1.1 Future User ...................................................... 50

4.1.2 The User story .................................................. 52
Final Documentation
4.1.3 Future Infrastructure ........................................ 52

4.2 Prototype timeline/ key learnings ..................................... 53
4.3 Finland convergence ....................................................... 53
4.3.1 Vision ............................................................... 56
4.3.2 Prototype features ............................................ 56
4.4 User testing ...................................................................... 56
4.5 Steering Wheel Concepts for Final Design ....................... 58
4.5.1 Magneto ........................................................... 58
4.5.2 Retracted steering wheel .................................. 59
4.6 Final Prototype Development ........................................... 60
4.6.1 Anticipatory Chair ............................................. 60
4.6.2 Interactive Steering Wheel ............................... 61
4.6.3 Open and Clear Cabin Space .......................... 61
5. Design Specifications .................................................................. 63
5.1 Anticipatory Chair ...........................................................
67
5.1.1 Chair Electronics ............................................
68
5.1.1.1 Microcontroller Board .....................
68
5.1.1.2 Chair Motor Driver Board ................
69
5.1.1.3 Rotation Motor Driver Board .............. 70

5.1.1.4 Chair Motor Sensing Board .............. 72
5.1.1.5 Rotation Optical Encoder ................. 73
5.1.1.6 Force Sensing Resistors .................. 74
5.1.1.7 Override Switches ............................ 74
5.1.1.8 Master Control Box .......................... 76
5.1.2 Firmware Development .................................... 77
5.1.2.1 Tilt Function
..................................... 78
5.1.2.2 Slide Forward/Back Function ............ 79

5.1.2.3. Rotate Function ............................... 79
5.1.2.4 Move Motor Function ....................... 80
5.1.3 Rotation Mechanism design ............................ 80
5.1.3.1 Assumptions ..................................... 81
10
Final Documentation
5.1.3.2 Design Calculations .......................... 81

5.1.3.3 Initial Rotation Prototype ................... 81
5.1.3.4 Final Rotation Mechanism ................. 82
5.1.4 Cabin Base Design ............................................. 83
5.1.5 Footpad Design ................................................. 84
5.2 Interactive Steering Wheel ................................................ 85
5.2.1 Interactive Steering Wheel Structure ................ 87
5.2.2 The Making of Interactive Steering Wheel ......... 88
5.2.3 The interaction specifications of the steering wheel 90
5.2.4 Comunication Between Devices ..............
95
5.3 Open and Clear Cabin Space .......................................... 95

5.3.1 The Making of Open and Clear Cabin Space ...
97
6. Project Planning and Management .............................................. 97

6.1 Timetable and Milestones ................................................ 103
6.2 Budget and Spendings ..................................................... 104
6.3 Distributed Team Management ......................................... 105
6.4 Stanford EXPE .................................................................. 111
6.5 Future Work ...................................................................... 112
6.6 Personal Reflections ......................................................... 113
7. Appendix .......................................................................................... 114
7.1 Initial Brief ......................................................................... 127
7.2 Fall Brochure .................................................................... 128
7.3 Needfinding ...................................................................... 129
7.4 Benchmarking .................................................................. 131
7.5 Technical Literature Benchmarking .................................. 132
7.6 Critical Function Prototype Golf Cart ............................
137
7.6.1 Golf Cart CFP Results .....................................
141
7.6.2 Golf Cart C# Code for Gradual Control ........
141
7.7 Online Research .............................................................. 143

7.8 Fall Presentations ............................................................. 154
7.9 Winter Presentation .......................................................... 157
Final Documentation
7.10 Winter Brochure
11
..............................................................
165
7.11 Dark Horse Prototype .......................................................
168
7.11.1 Reconfiguro - Drag and Drop Interface Java Code 170

7.12 Funky Prototype ................................................................. 170
7.12.1 Arduino Code for FSR ....................................... 191
7.12.2 Matlab Data Collection Code (main.m) .............
191
7.12.3 Matlab Data Collection Code (initialize_port.m)
192
7.12.4 Matlab Data Collection Code (readSerial.m) .....
193
7.12.5 Matlab Data Collection Code (writeToFile.m) ....
194
7.12.6 Matlab Post Process Code (post_proc.m) ........
194
7.13 Functional Prototype ......................................................... 194

7.13.1 Arduino Code - Anticipatory Chair ..................... 196
7.13.2 Arduino Code - Buckle it Out ..........................
196
7.14 Overview ........................................................................... 220

7.15 Design Reasoning ............................................................
226
7.16 Development Strategy ......................................................
227
7.17 Future Assumptions ..........................................................
229
7.17.1 Future user ........................................................
230
7.17.2 Future infrastructure ........................................... 231

7.17.3 Futute technology ............................................... 232
7.17.4 Ideal Future Persona ......................................... 233
7.18 Needfinding ........................................................................ 236
7.18.1 Context Map ....................................................... 238
7.18.2 Self-Observation ................................................
239
7.18.3 Interviews .......................................................... 239

7.18.4 EMT (Emergency Medical Technician) ..............
241
7.18.5 Survey ................................................................ 248

7.19 Benchmarking ................................................................... 249
7.19.1 Steering Benchmarking ..................................... 253
7.19.2 Motion Sickness Benchmark ............................. 254
7.19.3 Human-Machine Transition Benchmarking .......
259
12
Final Documentation
7.19.4 Confirmation Cue Benchmarking ...................... 260

7.19.5 Trust benchmarking ........................................ 262
7.20 Critical Prototypes .......................................................... 265
7.20.1 CFP - Transition Golf Cart ........................... 270
7.20.2 Steering Mechanisms .................................... 270
7.20.3 CEP - Reconfigurable Workspace ............... 275
7.20.4 CEP - Mobile Workspace ............................. 282
7.21 Design Specifications
.................................................... 284
7.21.1 Transition sequence prototype ....................... 286

7.21.2 Different Steering Controls ............................. 286
7.22 Needfinding .................................................................... 291
7.22.1 Trip to Germany ............................................. 296
7.22.2 Geneva Trip .................................................... 296
7.22.3 Dashboard Questionnaire .............................
298
7.23 Benchmarking ...............................................................
299
7.23.1 Akka Car Concept .........................................
301
7.24 Dark Horse Prototypes .....................................
301
7.24.1 Dark Horse Prototype - Reconfiguro ........... 302

7.24.2 Dark Horse Prototype - SbW Imitation ........ 302
7.24.3 Dark Horse Prototype - Sleeping Positions
310
7.24.4 Dark Horse Prototype - AR Imitation .........
312
7.24.5 DHP - Disappearing Steering Wheel ..........
314
7.24.6 Dark Horse Prototype - Magneto ................
315
7.25 Funky Prototypes ..........................................................
316
7.25.1 Funky Prototype Anticipatory Chair ............
318
7.25.2 Funky Prototype What should I do .............. 318

7.26 Functional System Prototypes ....................................... 325
7.26.1 FSP Anticipatory System ............................. 330
7.26.2 Functional System Prototype - Buckle It Out
330
Final Documentation
7.27 Design Specifications ....................................................
13
333
7.27.1 Design Specifications For Anticipatory Chair
336
7.27.2 Design Specifications For Anticipatory System
336
7.27.3 Design Specifications For Buckle it out ........ 341
14
Final Documentation
1.5 List of Figures

Figure 1.1.1: Final Prototype ............................................................... 5
Figure 2.3.1 - Audi Design Team ........................................................
27
Figure 2.3.1.1 - Stanford University Logo ..........................................
28
Figure 2.3.1.2 - Sangram Patil ...........................................................
28
Figure 2.3.1.3 - Stephanie Tomasetta ................................................
29
Figure 2.3.1.4 - David Wang ..............................................................
29
Figure 2.3.2.1 - Aalto University Logo ...............................................
30
Figure 2.3.2.2 - Goran Bjelajac ..........................................................
30
Figure 2.3.2.3 - Sifo Luo ....................................................................
31
Figure 2.3.2.4 - Heikki Sjman ..........................................................
31
Figure 2.3.2.5 - Tommi Tuulenmaki ...................................................
32
Figure 2.4.1 - Audi Logo ....................................................................
32
Figure 2.4.2 - ERL Logo ....................................................................
32
Figure 2.5.1 - Aalto Teaching Team ...................................................
34
Figure 4.1.2.1 - User story ..................................................................
52
Figure 4.2.1 - Steering Transition CFP ..............................................
53
Figure 4.2.2 - Steering Mechanisms .................................................
54
Figure 4.2.3 - Darkhorse prototype ...................................................
54
Figure 4.2.4 - Magneto prototype ...................................................
54
Figure 4.2.5 - Darkhorse prototype ...................................................
55
Figure 4.2.5 - Funky Prototype .........................................................
55
Figure 4.2.6 - Sensing chair ..............................................................
55
Figure 4.2.7 - Retractable steering wheel Aalto ................................
56
Figure 4.4.1 - User Testing ................................................................
58
Figure 4.4.2 - User Testing2 ..............................................................
58
Figure 4.5.1.1 - User Testing .............................................................
59
Figure 4.6.2.1 - Early Steering Wheel Mockup ..................................
62
Table 4.6.2.2 - Retraction Mechanism Explosion ..............................
63
Table 4.6.3.1 - Cabin Space mock up ...............................................
65
Figure 5.1.1.1 - System level view ....................................................
68
Final Documentation
15
Figure 5.1.1.2 - Physical layout of the hardware. ................................ 69

Figure 5.1.1.1.1 - Physical circuit board for the microcontroller ........... 70
Figure 5.1.1.2.1 - Schematics ............................................................. 71
Figure 5.1.1.2.2 - Physical circuit board .............................................
71
Figure 5.1.1.3.1 - Dual fan assembly for rotation motor heatsink .......
72
Figure 5.1.1.4.1 - Motor position sensing circuit schematic .................
73
Figure 5.1.1.4.2 - Motor position sensing board .................................
73
Figure 5.1.1.5.1 - Rotation encoder - optical sensor and circle of stripes 74

Figure 5.1.1.6.1 - Initial configuration of FSR tested ..........................
74
Figure 5.1.1.6.2 - FSR protoboard .....................................................
75
Figure 5.1.1.6.3 - FSR circuit schematic ............................................
76
Figure 5.1.1.7.1 - Switch override circuit schematic and protoboard
76
Figure 5.1.1.7.2 - Switch override protoboard ...................................
77
Figure 5.1.1.8.1 - Master control box ................................................
77
Figure 5.1.2.1 - Firmware process flow .............................................
78
Figure 5.1.3.2.1 - Chair Dimensions ..................................................
80
Figure 5.1.3.2.2 - Design Calculations ..............................................
81
Figure 5.1.3.3.1 - Rotation first prototype 1 ......................................
82
Figure 5.1.3.3.2 - Rotation first prototype 2 .......................................
82
Figure 5.1.3.4.1 - Rotation mechanism initial assembly ....................
83
Figure 5.1.3.4.2 - Rotation connection to platform .............................
83
Figure 5.1.3.5.3 - Rotation connection to motor ...............................
83
Figure 5.1.4.1 - Cabin base ...............................................................
84
Figure 5.1.4.2 - Drawing of the central cabin base design ...............
84
Figure 5.1.5.1 - top footpad layer ......................................................
85
Figure 5.1.5.2 - bottom footpad layer ................................................
85
Figure 5.1.5.3 - middle footpad layer ................................................
85
Figure 5.1.5.4 - Footpad top view showing wire routing ...................
85
Figure 5.1.5.5 - Footpad with embedded ball casters ......................
86
Figure 5.2.1 - 3d Printing of Steering Wheel ....................................
87
Figure 5.2.1.1 - Steering Wheel .......................................................
88
16
Final Documentation
Figure 5.2.1.2 - Steering Wheel Parts ................................................ 88

Figure 5.2.1.3 - Steering Wheel Parts Back
.....................................
89
Figure 5.2.1.4 - Retraction Mechanism Parts ....................................
89
Figure 5.2.2.1 - Base .........................................................................
91
Figure 5.2.2.2 - Base Dimensions .....................................................
91
Figure 5.2.2.3 - Tightening Bands .....................................................
92
Figure 5.2.2.4 - Tightening Bands Dimensions .................................
92
Figure 5.2.2.5 - Shaft ........................................................................
93
Figure 5.2.2.6 - Inner shaft technical drawing ..................................
94
Figure 5.2.2.7 - Microswitch support ................................................
94
Figure 5.2.2.8 - Assembled Retraction Mechanism ..........................
95
Figure 5.2.4.1 - Communication schematics ....................................
96
Figure 5.2.4.2 - Physical Implementation ..........................................
96
Figure 5.3.1.1 - Gluing the pieces .....................................................
97
Figure 5.3.1.2 - Sanding the dashboard ............................................
98
Figure 5.3.1.3 - Painted dashboard ...................................................
99
Figure 5.3.1.4 - Banding the ribbon ...................................................
100
Figure 5.3.1.5 - Barbecuing the ribbon ..............................................
100
Figure 5.3.1.6 - Mounting the ribbon ..................................................
101
Figure 6.4.1 - EXPE Booth ................................................................. 112
Final Documentation
17
1.6 List of Tables

Table 3.1.1 : Given Design Requirements .......................................... 36
Table 3.2.1 : Functional Requirements ............................................... 37
Table 3.2.1.1 : Functional Opportunities ............................................. 42
Table 3.3.1 : Physical Requirements ................................................... 43
Table 3.3.1.1: Physical Opportunities ................................................... 45
Table 3.4.1 : Business Opportunity ..................................................... 47
Table 4.3.2.1 Priorities for EXPE ........................................................ 57
Table 4.3.2.2 Priorities for EXPE ........................................................ 57
Table 4.5.1.1 Pro and Con for Magneto User Testing ........................
59
Table 4.5.2.1 Pro and Con for Retracting Steering Wheel .................
60
1.7 Glossary
Glossary
This system is much like regular cruise control, it is
Adaptive Cruise Control
supposed to maintain certain speed. Adaptive that the
ACC
car is able to adapt with current surroundings, it is able to

keep wanted distance with the car in front.
Adaptive Chair Reaction

(ACR)
The chair reaction to changes in body position of the user
Analog-to-Digital
This a device that converts a continuous physical quantity
Converter (ADC)
(eg. voltage) to a digital data.
18
Final Documentation
Glossary
Application Programming
This is a protocol that software components use as an
Interface (API)
interface for communicating with each other.
Arduino
This is a single-board microcontroller.
Augmented Reality (AR)
Autonomous Mode
Technology that combines actual world and digitally

generated image into one.
The mode when car is driving itself.
An exercise of exploring as many areas as possible that

Benchmarking
relate to the problem statement, such as technology

research, predictions, etc.
Cabin Space
The interior space within a vehicle which is occupied by

passengers/drivers.
This means information that the user of a car has to
Confirmation Clue
get about the events outside and inside the car. This
information confirms that car is behaving the way its user
wants it to behave.
Computer Numerically
Controlling method eg for machining tools that uses
Controlled (CNC)
parameters given by a computer.
Critical Experience
Physical prototype of an experience of design, which is
Prototype (CEP)
required to ensure usability.
Critical Function
Physical prototype of a fundamental element of design,
Prototype (CFP)
which is required to ensure its functionality.
Final Documentation
19
Glossary
Dashboard
Double Pole, Double
Throw (DPDT)
Electroencephalography
(EEG)
Control panel in front of the driver in a car.

A switch type that includes two SPDT switches
The recording of brains electrical activity along the scalp.
Electronically Erasable
Programmable Read-
Type of memory
Only Memory (EEPROM)

Emergency Medical
Term used in some countries to render to health care
Technician (EMT)
provider of an ambulance.
EXPE
Field Of View (FOV)
The presentation of project outcomes in the end of ME310

course at Stanford.
The vision field of Kinetic camera device that tracks body
movement
Force Sensing Resistor
This is a material whose resistance varies if a force is
(FSR)
applied.
Flowdoc is a collaboration web application for technical
Flowdock
team. It allows for quick updates, team communication,

and compiles all email threads in one place.
This is a term used for ME310 prototype that is still a bit
Functional System
Prototype
crude, and obviously assembled from off-the-shelf parts;

however, this time decisions on
technical implementation are done with increased
sophistication.
This is a term used in ME310 for an approximation
Funky Prototype
prototype of the full system without making a costly

commitment to any one configuration, technology, or
geometry.
20
Final Documentation
Glossary
H-Bridge
Haptics
Ideal Persona
Information Technology
(IT)
This is an electronic circuit that applies voltage across a

load in either direction.
Any form of nonverbal communication involving touch,
eg buttons and touch screens.
It is an assumption of potential extreme user in the
future.
The application of computers and telecommunications
equipment to store, retrieve, transmit and manipulate
data
In-car environment system, which combines audio
Infotainment System
on screen hardware for providing information and

entertainment.
Intravenous Therapy (IV
Therapy is the infusion of liquid substances directly into
Therapy)
a vein.
Infrared Camera (IR
Camera that detects not visible light with longer
Camera
wavelengths.
Input/Output Port (I/O
A port used in microcontrollers for two way
Port)
communication between a computer and microcontroller.
Integrated Drive
Electronics (IDE)
Kinetic
Kinetic
Lane Assistant
A standard for connecting a storage device-
Field Of View (FOV)

This is a motion sensing input device provided by
Microsoft.
This is a system which detects side lanes and tries to
keep the car in between them.
Final Documentation
21
Glossary
Mac OSX
Operating system of Apple computer devices.
Manual Mode
The mode when the user is controlling the car.
Matrix Laboratory
This is a numerical computing environment and fourth-
(MATLAB)
generation programming language.

This is a small computer on a single containing a
Microcontroller
processor core, memory, and programmable input/

output peripherals.
Mobile Workspace
Mode Initiators (MI)
Critical experience prototype that tested workspace

environment in real life traffic.
Commands that are used to transition between preset
modes.
It is sickness caused by difference between visually
Motion Sickness
perceived movement and the vestibular systems sense

of movement.
This is a device that selects one of several analog or
Multiplexer
digital input signals and forwards the selected input into

a single line.
An exercise of understanding and building empathy
Needfinding
for the target user group by conducting interviews and

ethnographic studies.
Night Vision
Original Equipment
Manufacturer (OEM)
This is a system that makes seeing possible in low light

conditions.
Standard component provided with the original product.
22
Final Documentation
Glossary
OpenNI/NITE
Physical Steering
Pull-Up Resistor
This is an open source API library.

Our critical function prototype that experimented
intuitiveness of different types of steering.
These are used in electrical circuits to make sure the
system settle at expected level.
Pulse-Width Modulation
This is a commonly used technique for controlling power
(PWM)
to inertial electrical devices.
Reconfigurable
Critical experience prototype of ours, which mimicked
Workspace
the transition between driving mode and leisure mode.

The angle of the chair at which the users start feeling
Sandwich Threshold
uncomfortable and are concerned that the chair is

sandwiching them.
Segway
Servo Motor
Two-wheeled self-balancing battery-powered electric

vehicle.
This is a rotary actuator that makes precise control of
angular position possible.
A library for Processing is a simple wrapper library that
SimpleOpen NI
maps simple call functions to the more complex API

library functions.
Situational Awareness
Skype
This means awareness that driver has about the

surroundings of the car.
Online communication tool which allow video calls and
chatting.
Single Pole, Double
A simple switch that changes connection of two
Throw (SPDT)
terminals to one common terminal.
Final Documentation
23
Glossary
Control system for an automobile that is done with
Steer-by-Wire (SbW)
electronic control systems using electromechanical

actuators and human-machine interfaces replacing the
mechanical control system.
Steering Wheel
Teaching Assistant (TA)
Teensy
Transition Golf Cart

Vehicle to Vehicle
communication (V2V)
Current round control device for a vehicle.

ME310 assistant that helps the students with their
project.
This is a single-board microcontroller
A critical function prototype which tested transition
between autonomous mode and manual mode
Communication between autonomous vehicles
Vehicle to Infrastructure
Communication between autonomous vehicles and
communication (V2I)
infrastructure.
VCC
Three letter combination used for power supply pin.
VDC
Volts in direct current.
Windows
Operating system for a personal computer provided by

Microsoft.
24
Final Documentation
Final Documentation
2. Context and Background
25
26
Final Documentation
2.1 Need Statement
adaptable in order to provide entertainment or
Audi has created a brand that people
mode, but also must provide a safe transition
respect and seek out because of their

utilization
of
advanced
technologies.
They are at the forefront of innovation in

automotive technology and are dedicated
to providing customers with elegant,
sophisticated solutions. The pursuit of new
technologies and with the future always in
mind, Audi has developed numerous driving
assistant systems and technologies. These
systems are bringing Audi one-step closer
to the implementation and introduction of
autonomous vehicles.
Autonomous vehicles will be relatively
common by the year 2035 and Audi envisions
car design will focus not only on the driving
experience, but also the riding experience.
As these technologies are adopted, people
will want to regain time lost from commuting
to locations where they would be productive.
Therefore, the interior cabin space design
must evolve from the current configuration
meant for driving to one that is able to adapt to
the various activities a driver and passenger
might want to do when in autonomous driving
mode.
2.2 Problem Statement

The goal of Team Audi Evolve is to develop
a cabin space that will allow Audi users,
in the year 2035, to perform numerous
activities in an autonomous vehicle. The
riding experience and driving experience of
an autonomous vehicle must be considered
in the solution. The cabin must be open and
workspace for the driver during autonomous

to manual driving when the user chooses to
take over the controls. The design solution
must ensure that the driver and passengers
trust the vehicles autonomous functions.
Furthermore, the solution must display all
necessary information for transitioning from
autonomous to manual mode and manual
to autonomous mode intuitively and without
being disruptive to the overall riding and
driving experience. The solution must also
have the sleek, sophisticated craftsmanship
that the Audi brand is known for and should
still retain the same look and feel of the brand.
2.3 The Design Team

The Audi 2012 design team is comprised
of a diverse and multi-disciplinary group of
students from Stanford University and Aalto
University who are excited to be working
together and learning from each other.
(Figure 2.3.1: Audi Design Team)
Final Documentation
Figure 2.3.1 : Audi Design Team (Thank you Djordje)
27
28
Final Documentation
2.3.1 The Stanford Team

Figure 2.3.1.1 : Stanford University Logo
Figure 2.3.1.2 : Sangram Patil
Stanford University
Sangram Patil
Location: Stanford, California (USA)
Status: 2nd Year Mechanical Engineering
Founded: 1891
Graduate Student
Contact: sangram@stanford.edu
Phone: 650.704.1145
After completing his undergrad in ME from
India, he joined Stanford last year and has
been enjoying an awesome roller-coaster
ride since then. Sangram thoroughly enjoys
working in team projects and being a part of
a student community that is so radiant and
full of enthusiasm.
Final Documentation
29
Figure 2.3.1.3 : Stephanie Tomasetta
Figure 2.3.1.4 : David Wang
Stephanie Tomasetta
David Wang
Status: 1st Year Mechanical Engineering
Status: 2nd Year Electrical Engineering
Graduate Student
Graduate Student
Contact: sltomase@stanford.edu
Contact: dcwang3@stanford.edu
Phone: 732.492.8373
Phone: 407.376.4635
Stephanie received her BSE in Product
Born and raised in Orlando, FL. David
Design from Stanford and decided she loved
received a BSEE from University of Florida
it so much she wanted to stay for a masters
(GO GATORS!). Interned at Lockheed
degree in Mechanical Engineering. She
Martin Missile and Fire Control in Orlando
believes design can bring delight to people
for 4 summers supporting both in-production
and hopes to design things that people will
and IRAD programs. As far as grad school,
become emotionally attached to, as well as
David is looking to graduate in the spring.
functionally.
Specialization/interest includes computer

architecture,
integration.
hardware
design
and
30
Final Documentation
2.3.2 The Aalto Team

Figure 2.3.2.1 : Aalto University Logo
Figure 2.3.2.2 : Goran Bjelajac
Aalto University
Goran Bjelajac
Location: Espoo, Finland
Status: Industrial and Strategic Design
Founded: 2010
Graduate Student
Contact: goran.bjelajac@aalto.fi
Phone: +385417009819
Goran Goci Bjelajac was born in 1986. in
Belgrade, Serbia. Since his early age he
has showed exceptional talent and interest
in art. A profession in design seemed natural
and the only option for him. He enrolled in
specialized high-school for industrial design
and has continued education at the Faculty
of applied arts in Belgrade and now at
Aalto University School of Art, Design and
Architecture at the department of Industrial
and Strategic Design.
Final Documentation
Figure 2.3.2.3 : Sifo Luo
Sifo Luo
Status: Information and Service
Management Graduate Student
Contact: sifo.luo@aalto.fi
Phone: +385417009819
Sifo got her Bsc in Business Technology
from Aalto School of Economics. She is now
continuing her Masters study in Information
and Service Management. Educated in three
countries -- China, Finland, and US, she is
very good at adaptation to various cultures.
Sense of urgency and teamwork are highly
valued by Sifo, and her new enthusiasm is
searching for inspiration in daily routine.
31
Figure 2.3.2.4 : Heikki Sjman
Heikki Sjman
Status: Mechanical Engineering Graduate
Student
Contact: heikki.sjoman@aalto.fi
Phone: +385417009819
Heikki is on his final year of his masters
degree in Mechatronics in Aalto University,
Finland. During his studies, he has been
working in the field of Design and spent a
year in United Kingdom studying business.
Heikki enjoys challenges and making things
happen. Every day is a new adventure!
32
Figure 2.3.2.5 : Tommi Tuulenmaki
Final Documentation
2.4 Corporate Sponsor
Figure 2.4.1 : Audi Logo
Tommi Tuulenmki
Status: Mechanical Engineering Graduate
Student
Contact: tommi.tuulenmaki@aalto.fi
Phone: +385417009819
Comes from the land of ice and snow,
Finland, where he did his undergrad in ME.
The journey continues, and now Tommi has
stepped on the path of a graduate student,
which is also being done at same school,
Aalto University. Tommi thinks challenges
are necessary for individual development;
this is why he also enjoys them so much.
He rejoices even more, when overcoming
the challenges is done with solid teamwork.
Figure 2.4.2 : ERL Logo
Audi is a subsidiary of the Volkswagen

Group, the largest automotive vehicle
manufacturer in 2011. The design team
is working closely with the Electronics
Research laboratory (ERL).
The ERL
strives to provide innovation and creativity

to the Volkswagen Group and all of its
brands. It is part of the global research
and development network that differentiates
its brands from the rest of the automotive
industry through its cutting edge technology
and research. The ERL performs research
and development in areas such as, human
machine interface systems, driver assist
systems, and infotainment applications and
platforms.
Final Documentation
33
VOLKSWAGEN Group of America, Inc.
Scott Steber, Teaching Assistant
Electronics Research Laboratory
Contact: steber@stanford.edu
500 Clipper Drive

Belmont, CA 94002
Jeremy Dabrowiak, Coach
Phone: +1 650.496.7000
Contact: jdabrowiak@gmail.com
2.4.1 Corporate Liaison

Trevor Shannon
Volkswagen Electronics Research Lab
Engineer, Multimedia Applications Team
Contact: trevor.shannon@vw.com
Phone (office): +1 650-496-7063
Phone (mobile):704.340.4160
Lorenz Bohrer
User Interaction designer
Contact: lorenz.bohrer@audi.de
Phone (office): +49 849 89 576866
Phone: 650.274.7871
Lauri Repokari, Professor
Contact: repokari@stanford.edu
Sara De Moitie, Teaching Assistant
Contact: sara.demoitie@gmail.com
Maria Kulse, Teaching Assistant
Contact: maria.kulse@gmail.com
Tuomas Sahramaa, Teaching Assistant
Contact: tuomas.sahramaa@gmail.com
Mikelis Studers, Teaching Assistant
Contact: mikelis.studers@aalto.fi
2.5 Teaching Team

Stanford University
Mark Cutkosky, Professor
Contact: cutkosky@stanford.edu
Larry Leifer, Professor
Contact: leifer@cdr.stanford.edu
George Toye, Professor
Contact: toye@cdr.stanford.edu
Tyler Bushnell, Teaching Assistant
Contact: busht@stanford.edu
Annika Matta, Teaching Assistant
Contact: amatta@stanford.edu
Harri Toivonen, Teaching Assistant

Contact: harri.k.toivonen@aalto.fi
Markku Kokela, Teaching Assistant
Contact: markku.koskela@aalto.fi
Lassi Laitinen, Coach
Contact: lassi.laitinen@gmail.com
2.6 Special Thanks to

George Atanasov, Marko Takala, Jordan
Davidson, Benjamin Tee
34
Figure 2.5.1 : Aalto Teaching Team
Final Documentation
Final Documentation
3. Design Requirements
35
36
Final Documentation
3.1 Given Requirements
solution. The functional requirements listed
Based on the benchmarking, needfinding
do and what functionality should be provided
and prototyping that was carried out in the

last six months, the team identified certain
basic functional and physical requirements
that the final solution should address.
The team also identified many interesting
directions and opportunities, which can be
explored while working towards a design
Design Requirement
in this section dictate what the system must

by the solution. The physical requirements
dictate what the system should be like
physically. Given requirements have been
listed based on the abstract that has been
provided. These requirements have either
been expanded and/or refined further based
on the testing done this quarter.
Rationale
In order for the driver to fully enjoy
Driver should trust cars autonomous functions
autonomous mode of the car, he/she should

have complete confidence in its control
Smooth and comfortable transitions between

modes and activities
The driver will perform a lot of activities in

autonomous mode and will need to transition
between them quickly and easily.
In order to appreciate freedom inside of the
Cabin should be clear and open
car, cabin has to be spacious enough and

clear so it provides comfortable working
environment.
Audi fit and finish

Table 3.1.1 : Given Design Requirements
The solution should have a similar level of

quality and usability as all Audi cars
Table 3.2.1 : Functional Requirements Part1
achieved the interface for this

transition which will reconfigure
the cabin space needs to blend in
reconfiguration interface 3 times,
they should be comfortable using
it at any point later.
current mode of the configuration
at all times during transition. If
asked about the current mode at
any point during transition, the
user should be able to identify it
used should
indicate to the
user that the
transition is
complete.
want to perform when in autonomous

mode can be performed while being
in the driving position.
or orientation. The design should
allow the user to move a distance
which is at least 2 times the width
should have
the flexibility of
moving around
of the seating or chair
Not all activities that the user would
mode, the interface is locked,
reassurance of trust on the system.
and be comfortable at any angle
input device for working; in manual
mode of the car will also lead to a
and sit in any orientation.
more mobility and be able to move
chair allows the driver to have much
The rotation mechanism within the
showing an image of the Audi logo.
is a unlocked so it can be used as an
different
Being kept aware of the current
two
In autonomous mode, the interface
has
and comfort during the transition.
tablet
interfaces for two different modes.
The
she is on.
confirmation on which mode he/
The driver will receive clear
learn interaction with the chair.
been changed, and an easy-to-
an audible click when mode has
through having a tablet that has
This requirement is achieved
Implementation
This is required for user acceptability
The cabin space
The user should be able to turn
correctly at any time.
amongst the users in terms of the
There should be 100% awareness
activities. For this goal to be
various interactions with the
with the transition itself.
comfortable transition between
The team wants to target a
Rationale
After new users go through the
Metric
The interface
user.
intuitive to the
space should be
of the cabin
reconfiguration
The interface for
Functional
Requirements
Final Documentation
37
3.2 Functional Requirements
and activities
between modes
transitions
comfortable
It makes sense to provide

maximum functionality with
minimum clutter in terms of the
controls.
The controller must be able to
perform at least 5 activities that
would want to be performed in
the vehicle, including driving.
home.
around the driver
Smooth and
more spacious like a room in a
cabin space
vehicle and instead needs to be
volumetric
passengers.
other than drive.
utilization of
F2F interaction with backseat
will have free time to do things
changes.
into the next activities position.
allows the user to naturally flow
tedious manual adjustments and
which removes the need for
chair is an adaptive interaction
user to perform in any mode. The
is proven easy and intuitive for
wheel done by pushing and pulling
The mode trigger on the steering
free and spacious.
seat, the cabin space is now more
By removing the front passenger
improved socialization through
since in autonomous mode they
space through minimal cabin
feel like a tight transportation
chair brings more flexibility and
for a number of different activities
easily be performed in the cabin
be maximum
through the tablet interface; the
for passengers to use the space
different activities that can
The cabin space should no longer
to perform various activities
The cabin allows the driver
This is for developing an

adaptable cabin space that allows
Implementation
Rationale
There should be at least 5
Metric
There should
mode
in autonomous
multiple activities
performing the
well suited for
Cabin has to be
Functional
Requirements
38
Final Documentation
retraction mechanism is fast to

activate.
time the user wants to do a

different activity
10 that measures the sporty
feeling of an Audi control input
and sporty
feeling of driving
driving
autonomous
of monitoring
responsibility
driver of the
Relieves the
in every 3 minutes
information not more than once
User checks or verifies the car
to be 7 or higher on a scale of
precise control
an Audi
score the control input response
maintain the
and enjoy doing other activities.
will help users focus better on
the autonomous controller actions
Reducing the distraction due to
of controlling a complex machine
of driving is the feeling and thrill
factor contributing to the pleasure
be active for safety. An important
proof systems, which will always
redefined due to the added crash
manual driving has been
on the windshield.
used for controlling work contents
in autonomous mode and can be
The steering wheel can not move
driving experience.
car to further enhance a sporty
mimics the setup of a formula race
the centered position of the chair
increases the sporty feeling, while
next position.The steering wheel
reconfiguring the space every
This is an essential requirement
quickly move the driver into their
associated with manually
A responsible driver should
seconds
and activities
to rotate at a comfortable speed to
at reducing the tedious tasks
design should
next activity is less than 15
between modes
The chair mechanism is designed
The smart cabin space aims
for cars of the future, where
The time it takes to begin the
Quick transitions
Implementation
Rationale
Control input
Metric
Functional
Requirements
Final Documentation
39
should not go below 90
degrees.
threshold in
terms of the
the user
comfortable for
needs to
steering wheel
time of the
disrupts the smooth flow between
pushing or pulling.
activities.
like something is broken and it
If it takes longer for the steering
response
respond immediately after
chair response
too sensitive
without triggering the chair
wheel to respond, users feel
times should not trigger the
should not be
adjustments to their positions
Users must be able make minor
might get stuck in the chair.
and become confused that they
users get sandwiched in the chair
The steering wheel needs to
position by 20 degrees multiple
body commands
The emergence
Shifting or tilting upper body
The intentional
position
and the bottom of the chair
activities
space
the sandwich
be an interruption to the flow of
foot to reconfigure the cabin
If the angle is too low, then the
should be easy to use and not
move their arms more than a
The angle between the back
The mode initiator command
Rationale
The user does not need to
Metric
not exceed
The chair should
of the user
be within reach
command should
initiator
Any mode
Functional
Requirements
immediately after being initiated.
The steering wheel will respond
times in a row.
the command to happen multiple
secs) in software is to prevent
The time threshold (0.7-1.5
distrustful of the system.
angle so the user dont become
that the chair can never pass this
The chair has been coded so
facing forward.
of range as long as the driver is
wont make the steering wheel out
The steering wheel retraction
Implementation
40
Final Documentation
communication with back seat

passengers. When driving, the
only rotate fully back but also
move backward and forward.
to be moved or activated with
smart features in manual mode.
off when in
manual driving
mode.
during manual driving mode no
The chair should not be able
must be turned
unsafe.
he/she performs as it would be
matter what body movement
The driver shouldnt be rotated
driving position.
driver should be fit into the best
backseat to achieve better F2F
rotate back and move closer to
The driver should be able to
Rationale
The chair should be able to not
Metric
Smart features
the user
comfortable for
needs to
steering wheel
time of the
The emergence
Functional
Requirements
disabled.
the chairs smart features are
the manual mode is started
The chair is coded so that when
environment is maintained.
a comfortable communication
passengers is 40 inches, so that
btw foot pad and the backseat
to the back and the distance
The chair can rotate 180 degrees
Implementation
Final Documentation
41
Table 3.2.1.1 : Functional Opportunities
autonomous mode windshield will be used
multimedia while in autonomous mode.
using the same car. The cabin space needs

to be smart enough to readjust according to
preferences without any input from the
driver
the latest technological solutions in their cars.
technologies quickly
model like the current smart phones.
It can even be an upgrade based business
Audis goal to provide their customers with
Cabin should be able to integrate new
technology development in the future and
This is to be in line with the fast paced
the user.
scenarios where different drivers would be
This is especially applicable in car sharing
contents in manual driving mode, and in
the windshield can be used to interact with
For example: AR windshield
and adjust all settings to that specified user.
recognize different default sitting positions
It could be possible for the car to begin to
as working interface.
info will be shown on the windshield as AR
In our concept, speedometer and driving
Implementation
Since the users are not driving anymore,
Rationale
Should be able to recognize driver
surface
Using the windshield as an interactive
Opportunity
42
Final Documentation
3.2.1 Functional Opportunities
Table 3.3.1 : Physical Requirements
with brand AUDI
design must be in line
The fit and finish of the
use
and be comfortable to
the pleasure of driving
inputs should maintain
The designed control
be open and clear
The cabin space should
Physical Requirements
all
integrated
must
experience
with 85% accuracy.
prototypes manufacturer
able to identify Audi as the
Car enthusiasts must be
of the car on the road.
maintain feeling their control
testing survey), in order to
real (8 out of 10 on a post
mode that they deem to be
control input when in driving
haptic feedback from the
Driver
out without hitting anything.
able to stretch their legs
four passengers must be
features are stowed away,
When
Metric
appealing to the Audi user.
Audis in order to be consistent and
finish of the design must align with
fine styling. Therefore, the fit and
sophistication achieved through their
Audis brand is its refinement and
Users are brand loyal and part of
will want to do, not have to do.
since driving is an activity the driver
they should be easy to use and fun
longer need to be precise. Instead,
The driving control inputs may not
relax, or socialize.
to have the room to be able to work,
more comfortable in order for people
The cabin space must be open and
Rationale
to the Audi tradition.
2035. The color scheme remains true
believes Audi will look like in the year
The aesthetics fit into what the team
seat.
the removal of the front passenger
The cabin is much more open due to
in the backseat as well as the driver.
There are seats for three passengers
Implementation
Final Documentation
43
3.3 Physical Requirements
Table 3.3.1: Physical Requirements
tell the difference between
a smart chair and a normal
car seat in terms of comfort
The driver shouldnt put
more than 5 pounds of
force on food pad.
be concealed and
unnoticeable to the
user sitting in it
should
The
input minimal effort to
press the foot pad.
Users should be able to
engage the footpad with
be suitable size and
not slippery.
There should be atleast 1
of clearance between any
point on the chair and the
furthermost edge of the
steering wheel during the
entire rotation cycle.
The chair should clear
the steering wheel and
the dashboard while
rotating while being in
a comfortable recline
position for the user
flat bottomed shoes.
backseat passengers feet.
The foot pad should
The footpad should not hit
Users should not be able to
Force sensors must
driver
Metric
Physical Requirements
to
ensure
contact
with
able to engage the footpad with flat

bottomed shoes.
able to engage the footpad with flat

bottomed shoes.
rotation without hitting anything
To allow for a complete uninterrupted
passengers feet. Users should be
passengers feet. Users should be
rotation.
it might hit the steering wheel during
chair to this clearance position before
for the users. Software brings the
degrees of incline which is comfortable
that it will clear the steering wheel at 30
The axis of rotation was placed such
The footpad should not hit backseat
allow the plate to move easily
embedded FSRs and ball bearings to
points
The footpad has integrated raised
covering.
to the foam seat below the leather
The force sensors have been attached
Implementation
The footpad should not hit backseat
linear movement.
should be enough to activate chairs
driver. A minimum force of input
and effortless as possible for the
The transition should be as natural
sacrifice basic user comfort in a car
Adding smart features should not
Rationale
44
Final Documentation
Table 3.3.1.1: Physical Opportunities
being constrained to the internal structure of current cars
If there is personal ownership of the car in the future and the cabin
space has been so well designed it might as well integrate into the
home space instead of being parked away in a garage
structure itself
Having the car space
as a part of the users
home
using
Virtual
space
Reality in the cabin
of
Exploring the options
goes beyond the constraints of cabin spaces.
reality helmet and transport the users wherever they want to be. This
space. One direction of the solution might just be to have a virtual
might not be necessary to have the complete functionality in the cabin
VR technology would probably be so well developed in the future that it
lives of people
The car transforms into a personal robot to better integrate into the
This opportunity allows for more flexibility for the cabin space without
Redesigning the car
Transformer car
Rationale
Opportunity
by the current car constraints.
rotation feature and not be restricted
in order to incorporate the new
The prototype is built on a platform
Implementation
Final Documentation
45
3.3.1 Physical Opportunities
inside
Table 3.3.1.1.: Physical Opportunities Part 2
cabin
reconfigurable
space should be
The
interactive inputs
the cabin space as
windows
Utilizing walls and
away, but still within the users reach (2.5ft) at any chair location.
The cabin must have integrated tables and storage that are stowed
of cabin space.
This is one direction, which can be explored for maximum utilization
more fun while being easier to transition to.
into a more intuitive
control input
Designing a new control input like this has the potential to make driving
Rationale
steering and pedals
Combining both the
Opportunity
is
currently
only
being
suited for that activity.
away items in the space to be better
rearrange, and introduce or stow
activities so users may want to
The cabin will be used for various
the car.
implemented on the windshield of
This
Implementation
46
Final Documentation
Final Documentation
47
3.4 Business Opportunity

Opportunity
Rationale
After the launch of autonomous cars, the
business model can shift to car sharing or a
New business model for AUDI
subscription based model which will redefine

ownership of cars and urban mobility in the
future in general.
Table 3.4.1 : Business Opportunity
48
Final Documentation
Final Documentation
4. Design Development
49
50
Final Documentation
Audi is a subsidiary of the Volkswagen
of the future are well integrated in the lives
Group, the largest automotive vehicle
of users, the team can envision that people
manufacturer in 2011. The design team
would show the same enthusiasm towards
is working closely with the Electronics
these autonomous machines and new
Research laboratory (ERL).
upcoming technology as they are currently
The ERL
strives to provide innovation and creativity
showing towards smartphones.
to the Volkswagen Group and all of its

brands. It is part of the global research
Internet
and development network that differentiates
surpassed boundaries of countries in terms
its brands from the rest of the automotive
of connecting people across distant areas
industry through its cutting edge technology
of the world. Accessing a wide range of
and research. The ERL performs research
information has never been this easy and
and development in areas such as, human
it is going to get better in the future. There
machine interface systems, driver assist
is this concept of perpetual connectivity
systems, and infotainment applications and
which is being predicted for the future. It is
platforms.
only a matter of time before all the existing

Assumptions help the team to build a picture
of what the world will be in the future. In
order to get a feasible frame for the design
work, especially when it involved designing
something so far in the future, the team
made reasonable assumptions based
on basic trends from the past to present,
research, needfinding and benchmarking.
Sources for these iterations can be found
among appendices.
4.1.1 Future User
The current generation has grown up in the
technology revolution. People are use to
the fast paced development in technology
and the effect of these new technologies
on our lives is continually growing.
Smartphones and smart handheld devices
have redefined the way we interact with
digital media. If smart and self-aware cars
and
social
networks
have
technologies are developed to such an extent

that they become a part of our lives and
blend in so well that we cannot do without
them. Specifically in terms of automotive
experiences, it is already being seen that
the current generation views driving as a
distraction from texting rather than the other
way around. Since being connected is so
easy, people want to stay connected. This
does not mean that the future users would
not love driving.
Experiencing
moments
of
thrill
and
adventure would still be desirable in the

future, but the essential difference would
lie in doing things because users want to
do them rather than spending time being
forced to do them. It is an extrapolation of
the current scenario when people are forced
to drive along freeways with all the traffic
only because they have to travel to and
from work everyday. Based on this insight,
Final Documentation
it is highly probable that users will find it

desirable to have an option to ride in a cabin
space that is customized to their needs and
the activities that they would want to do, to
better utilize this time lost in commuting.
The internet age has also led to more liberal
thinking. Non-conventional work options
are being explored. It has been envisioned
that new technology will lead to a great
shift in working spaces, work cultures and
procedures. The future users are most likely
going to work in an environment where
physical presence is no longer required on
a daily basis. In such a situation and with the
increasing influence of autonomous cars, it
is highly probable that the future user is not a
very good driver without the basic assistance
systems. The perceived completely manual
mode of driving is very different from the
existing perception of manual driving. There
will be many assistance systems in place
in the future cars. This prediction can be
justified on the basis of experiences of pilots
in airlines which were fit with autopilots and
new assistance systems in the 20th century.
There was a time when all the pilots were
skeptical about adopting, getting used to
and trusting this technology. But currently,
pilots rely so much on this new technology
that most of them cannot do without it.
51
52
Final Documentation
4.1.2 The User Story
who helps his son prepare for a spelling
In thinking about the future Audi user, the
test, but also enjoys playing with his son
team envisioned a scenario in which the
and making him laugh. This means that
typical user would experience their car on
often he will go into driving mode to have
a daily basis. The team sees this user as a
fun and spend some quality time not only
businessman who spent his 20s and 30s
studying, but also sharing a bond over the
very focused on forwarding his career. He
pleasure of driving an awesome car. Once
gets married later in life and decides to
his son is safely at school, the Audi user is
have a child in his early forties. He enjoys
off to work and decides to send some last
taking his wife out on nice dinner dates,
minute emails to confirm several business
but also enjoys entertaining clients and
appoints and then relaxes until he reaches
going out with his buddies.
his office. This user story helped guide in

making our decisions about the prototype
This Audi user is a family man who typi-
and what features were most important.
cally drops his young son off at school
The figures below show visualizations of
on his way to work. He is an active father
this user story
Figure 4.1.2.1 : User story
Final Documentation
4.1.3 The Future Infrastructure
Number of cars will radically increase

globally
Number of cars in developed countries will decrease and in developing

countries will raise
Developed countries will shift towards

electric vehicles, but in developing
countries gasoline will still be the
primary fuel
China will be the biggest market for

cars. It will also be the largest car
manufacturer in the world
Middle east will lose its position in oil

based wealth, because of the lack
53
4.2 Prototype timeline/

key learnings
CFP - Steering Transition
QUESTION - Do people feel safe and
comfortable making the transition using
this prototype and does this setup increase the situational awareness for the
driver?
Insight/Conclusion - the interactive interface was too distracting and users were
better at a direct transition than a gradual
transition. (Figure 4.2.1: Steering Transition CFP)
of interest for oil and because the oil

reserves are going to expire
Because the number of people will

rise on the planet, there will be huge
demand for food thus countries that
are rich in farmland will be major powers in the future
Because of global warming, ice in

Siberia and north of Canada and
Europe will melt, which will make
those countries new world leaders in
farmland and as such the new global
leaders
By 2030, transition between real and

virtual world will be complete. User
will feel and see with senses everything what a person in virtual world
sees and feels. The technology will be
already available by 2020, but it will
not be safe and legal until 2030
More and more people will work from

home and live with parents - which will
bring a major decline in marriages.
Figure 4.2.1 : Steering Transition CFP
54
Final Documentation
CFP - Steering Mechanisms

QUESTION - Do users feel more comfortable using steering mechanisms other
than the traditional steering wheel for
driving?
Insight/Conclusion- Most users felt uncomfortable using other steering mechanisms as control inputs to driving since
they were trained with and always use the
traditional steering wheel.
Darkhorse Prototype - Reconfigurable
Workspace
QUESTIONS - Do users value a reconfig-
Figure 4.2.2 : Steering Mechanisms
urable cabin space in a car? What level

of control do users want over the cabin
space layout?
Insight/Conclusion - Users value the
benefits of being able to optimize the
cabin space for different activities but they
would like to do so in the least amount of
effort and intrusion in their daily lives. Users also perform a lot of activities quickly
and can not plan in advance what they
want to do. (Figure 4.2.3)
Darkhorse Prototype - Magneto
Figure 4.2.3 : Darkhorse prototype
QUESTIONS - What would the users do

with the steering wheel if it was possible to
attach it anywhere on the dashboard?
Insight/Conclusion - Users felt that the
wheel could be used as an interactive or
control device while in autonomous mode.
However the detachable wheel was cumbersome and users didnt know where to
place it. (Figure 4.2.4 Magneto)
Figure 4.2.4 : Magneto prototype
Final Documentation
55
Funky Prototype - Chair Sense

QUESTIONS - Is having a chair anticipate
the users desire to adjust the chair valued? Can force sensors be used to predict
intentions based on body language and
position of in a chair?
Insight/Conclusion - Users had a
positive experience with this anticipatory
sensing chair prototype when the chair
matched their intentions. Force sensors
were able to identify simple user intentions
like pushing back on the chair. Users
felt the chair made transitioning between
activities easier. (Figure 4.2.5)
Figure 4.2.5 : Darkhorse prototype
Funky Prototype - Methods of Transitioning Modes

QUESTIONS - Are their any intuitive
methods of transitioning between driving
and autonomous mode?
Insight/Conclusion - Voice recognition
and touchable buttons seemed to be the
most intuitive with hand gestures as an
alternative to initiate modes. The team observed users consistently pulling or pushing the steering wheel to make it come out
or retract. (Figure 4.2.6)
Figure 4.2.5 : Funky Prototype
Functional Prototype - Chair Sense 2.0

QUESTIONS - How well can people adapt
to this new chair interaction? Are gestures
an appropriate method to initiate transition
between driving and autonomous mode?
Insight/Conclusion - People enjoyed the
experience and concept of the prototype.
Gestures used as mode initiators were unreliable due to random misfires. Must accommodate a variety of user body types.
Figure 4.2.6 : Sensing chair
56
Final Documentation
Figure 4.2.7 retractable steering wheel aalto
4.3.1 Vision
The team was split in terms of whether to
focus on the activities or transitions between activities. A resolution was reached
and encapsulated in the vision. The focus
of the project was to facilitate and make
sure that users would be able to perform
many activities, while still making sure that
the transitions would be made much more
effortless. The team agreed to deliver a
complete experience so that users could
get a taste of what it would be like to work,
relax, and socialize in the cabin space
Functional Prototype - Retractable
Steering Wheel
QUESTIONS - Do users feel that initiating
modes with the seat belt is intuitive?
Insight/Conclusion - Seat belt as a trigger
brought confusion. The trigger has to be
something simple, clear, and notify the
user of which mode they are in.
redesign. This experience also included a

very intuitive and effortless way of moving through these three activities without
any tedious adjustments or extra steps,
as well as making sure the space was
flexible enough to allow for large reconfigurations, such as the rotation that would
be needed in order to improve socializing.
The results vision statement is as follows:
(Figure 4.2.7 retractable steering wheel

aalto)
4.3 Finland convergence

During the Stanfords team visit to Finland,
the whole team unified their vision, decided on what features were more important,
To facilitate activities
for the driver through
effortless transitions
and increased mobility.
divided tasks, and prioritized tasks. This

section will outline the final outcome of the
weeks discussions and work.
4.3.2 Prototype features

On the right page is a table of key elements of the final prototype that the team
wanted to integrate. It also outlines what
team was taking on the responsibility and
task, as well as the priority level of each
task.
Final Documentation
57
Table 4.3.2.1 Priorities for EXPE
Prototype Component
Team Division
Priority Level
Anticipatory Chair with sensors
Stanford
Very High
Rotation Mechanism
Stanford
Very High
Foot Pad to be integrated with chair
Stanford
Mid
Mock Cabin Design & Build
Aalto & Stanford
High
Retraction Mechanism
Aalto
Very High
Steering Wheel Design
Aalto
Mid
Interactive Steering Input Device
Aalto
Low
Windshield Driving Video/ Simulator
Stanford
Low
Manual Mode indicator
Aalto
High
Prototype Component
Team Division
Priority Level
Kinect- secondary cues for chair
Stanford
Very High
Butterfly Doors
Aalto
Low
Temperature Control
Stanford
Low
Sounds and lights
Aalto
Mid
Lighter chairs
Stanford
Md
Interactive Display
Aalto
Very high
Other features that the team identified as

nice to have and prioritization of these
tasks, as well as who would spearhead
the efforts if time permitted.
Table 4.3.2.2 Priorities for EXPE
58
Final Documentation
4.4 User testing

Throughout the build process, the team
tested with various users. The hope was
to test the chair with as many varying
body types as possible, in order to make
sure that the chair could be used comfortable with a wide variety of people. The following are some key points and discoveries that the team found during our various
testing rounds:
(Figure 4.4.1)
Chair Findings:
Short users were unable to reach the

floor to use the foot pad if sitting all
the way back in the chair
It took users a couple of minutes to

get comfortable with using the chair.
Figure 4.4.1 - User Testing
Many users tried pushing harder, even

though the FSRs were able to pick up
on much more subtle movements.
One user tried avoiding putting his

feet on the foot pad
Many taller users who had broader

shoulders had to lift their shoulder off
the chair in order to get the recline
adjustment to activate.
People had a hard time activating

rotation when asked to twist their bodies. But they found a much easier way
by just sliding their upper body in the
direction they wanted to rotate.
Users were delighted when they activated the rotation and began smiling
or giggling.
(Figure 4.4.2)
Steering Wheel Findings:
Users thought the clicking sound and

feel was gratifying when pulling the
steering wheel
Pulling the steering wheel took a little

too much effort.
Having the same background for both

modes made people unsure of what
mode it was in, even though it said the
mode on the screen.
Figure 4.4.2 - User Testing2
4.5 Steering Wheel

Concepts for Final Design
The team decided to test different concepts of a retracted steering wheel
mechanism to find out what is the most
intuitive way for the driver to switch to/
from different driving modes. The trigger
Final Documentation
59
Figure 4.5.1.1 - User Testing
should be easy to reach, fast to initiate,

and be able to integrate with the function
of the chair.
4.5.1
Magneto
The team took the magneto darkhorse

idea and refined the concept. The new
magneto steering wheel contained a
magnet on the central back area. The
magneto steering wheel was shaped into
a rectangular object since it was formed
around a tablet device. It was integrated
with a ribbon shaped dashboard made of
metal. To change from manual to autonomous mode, the driver could simply take
the steering wheel off from the dashboard,
and use it as an input device for working wheel back in the correct spot on the
User testing result:
Pros
dashboard. (Figure 4.5.1.1) ing, or attach

it to the extended part of the dashboard
or on the lower area of the chair which
were also made of metal. To change from
autonomous to manual mode, the driver
had to put the steer back in the steering
position.
Cons
Users put the magneto back on the driving
Fast and easy to change modes
spot even though they didnt want to drive,

because that was the best reading position
for the tablet device.
Clear confirmation of each mode
Users put the magneto back on driving spot

but did not realize they were in driving mode.
Users didnt know where to store the
Users loved the new concept
magneto/tablet after taking it off, confused

at first what to do with it.
Users were comfortable and relaxed while
Users felt unsafe to take the steering wheel
using the steering wheel as an input device
off
Users passed the magneto to the back seat

passenger to share the experience
Users loved the open view brought by the
ribbon like dashboard.
Table 4.5.1.1 Pro and Con for Magneto User Testing
60
Final Documentation
was triggered by a linear actuator. The
4.6 Final Prototype

Development
shape of the steering wheel maintained
AudiEvolve is comprised of three main
the same shape as conventional wheels.
components: an anticipatory chair that
To change from manual to autonomous
senses a users intentions from intuitive
mode, the driver has to apply a certain
body movements and adapts the chairs
pushing force to the steering wheel for
position accordingly, an interactive steer-
3 seconds. The wheel will then start to
ing wheel with touch screen display that
retract and blend into the dashboard as a
is used not only for transitioning between
flat surface. To change from autonomous
autonomous and driving mode but for in-
mode, the driver had to apply another 3
teracting with any digital content the user
seconds push on the steering wheel to
desires, and an open/clear cabin space to
initiate the steering wheel to come out.
maximize utilization of space for the user.
4.5.2
Retracted steering wheel
The retracted steering wheel mechanism
User testing result:

Pros
Users felt safe and reliable with little change

from legacy
Push movement is really intuitive
Cons
The driving position is not ergonomic
Mode change process is too slow
Users used a lot of force and waited for too

long
Users had to push the wheel back with both
hands, which is an unsafe position if you are
really driving
Pushing added force on the chair that
moved the chair back
Table 4.5.2.1 Pro and Con for Retracting Steering Wheel
Final Documentation
61
4.6.1 Anticipatory Chair
between activities and modes is effortless.
The final design of the anticipatory chair
The team designed the chair system to
was built upon the concept from the
have three main functions: a tilt, a slide,
teams functional prototype. The team
and a rotate function. Users wanted
focused on the driver and the transition
the ability to move freely without being
between activities that the driver would
disrupted from what they were doing and
most likely perform within the cabin space
using body movements that correspond to
of an autonomous vehicle. It was evident
how the user would move around in a roll-
that the chair is an important aspect within
ing/swivel chair was reasonable.
this design space. The chair has a variety

of pressure sensors embedded into it that
4.6.2 Interactive Steering Wheel
determines how the user wants to adjust
To give users a futuristic feeling, the team
the chairs position. Several major differ-
wanted to redesign the input device for
ences from the functional prototype to the
steering. CFP Physical Steering showed
final design was the modification of the
the team that a wheel like steering device
rotation mechanism to allow the user to
will be the best choice. The team wanted
smoothly rotate up to 180 degrees and
to maintain some legacy from old cars and
the addition of a footpad to allow the user
spice it up with a modern design.
to slide the chair back and forth. Rotating the user up to 180 degrees drastically
The team also added more interface value
changes socialization and interaction in
for the user within the steering wheel. The
the car with fellow passengers compared
darkhorse prototype Magneto showed us
to the conventional cabin space layouts.
that there are many additional opportuni-
The footpad also serves as a resting place
ties for the steering wheel since the car is
for the users feet, while the chair is rotat-
now autonomous. From the needfinding,
ing so the users feet do not drag across
benchmarking, user testing, and online
the floor.
research, the team discovered that the

location of the steering wheel is also the
From the functional prototype, the team
most optimal location for digital interac-
realized that the user experience and
tion in autonomous mode. Therefore the
operation of the system must be flawless
team designed a steering wheel, which
or the credibility of the system is compro-
incorporated an interactive surface on it.
mised. Therefore, the team implemented
Figure 4.6.2.1 shows early mock up of the
a manual override, as well as making sure
steering wheel design.
that the chair would never be capable of
There are five main key insights that af-
sandwiching a user in the chair. Ease of
fected the design solution:
usability and flexibility are important aspects for the user to feel that the transition
62
Final Documentation
Figure 4.6.2.1 - Early Steering Wheel Mockup
1. Interaction with the surface of the
The team designed an interactive surface
steering wheel has increased since
on the wheel for increased interaction
it was introduced as a steering input
possibilities and a retraction mechanism
method. First it was only for steering,
that allowed mode triggering with push
but nowadays it has all kinds of but-
and pull movements effortlessly. The
tons on it, so that the user is allowed
retraction mechanism also locks the steer-
to interact with it. Since future cars are
ing wheel while in autonomous mode and
autonomous it allows us to increase
aligns it with the dashboard.
this interaction even more.
2. The surface of the steering wheel is
To make triggering effortless, the team
the most accessible when inside the
designed a spring retraction mechanism.
cabin space for the driver.
The strength of the spring had to be
3. Users did not like to see the steering
strong enough to keep the steering wheel
wheel move completely to a different
pushed out but also weak enough so
area.
that the user can easily push the steering
4. Physically triggered modes make the
wheel in. Additionally, the steering wheel
mode change more apparent for the
had to be able to stay locked in when
user.
pushed. Quick calculations showed us
5. Push and pull triggers were most

natural ways for triggering modes.
that the spring had to be about 40 newtons. This meant that around 4kg weight
can push the steering wheel in.
Final Documentation
63
Table 4.6.2.2 Retraction Mechanism Explosion
4.6.3 Open and Clear Cabin Space
conducted research about the future user,
It is important that the cabin space lay-
their needs, and explored the possibilities
out was open and clear. In this design,
that future cars would bring. If the car is
there is only one car seat in the front,
autonomous, the driver does not have to
arranged in the middle of the car. The old
watch the road all the time. This change
fashioned dashboard was also removed
radically transforms socialization inside
since the team discovered electric cars
of the vehicle. As the teams focus is on
will dominate the car market, therefore
the transition between multiple activities,
there would be no need for large engine
the driver requires increased mobility
compartments. Steer by wire will be the
inside of the cabin. The front passenger
main technology for future steering wheel
seat was eliminated and the driver seat
connections allowing the windshield to
was positioned in the middle to drastically
be more prominent so passengers have
increase mobility and flexibility within the
a better view of the driving scene. With
cabin space. This allowed the front seat
these changes, the team dramatically
to be able to rotate 360 degrees, making
increased the openness and clearness
socialization between passengers more
within the cabin space.
natural and comfortable.

Though the team focused on the transition
In order to better understand what the
between activities, it was still important
cabin space should look like, the team
to understand what positions would be
64
Final Documentation
needed for the most common activities
inside of a car as well. In order to do so, it
that the team identified in a car, which
was necessary to have a front seat rotate
included: driving, working, socializing, and
at least 180 degrees.
relaxing.
Relaxing
Driving
Relaxing will be done with the seat and
The team spent the beginning of the year
position adjustments. When the car is in
researching different driving methods.
autonomous mode, the driver has to feel
After user testing multiple prototypes and
comfortable and be able to freely move
many interviews, it was clear that the
around.
steering wheel is still the best way to steer

the car.
The final cabin space design utilizes all of

the advantages that an autonomous car
Working / Entertainment
offers. The team wanted to create some-
There are multiple prototypes and emerg-
thing that will give the a feeling of being in
ing technologies which have the potential
the car and communicate the team vision
to radically change the way people inter-
in the best possible way.
act with machines and work. Considering

current trends, work in the future will be
The designed cabin is open and clear.
digital and people will become even more
There is no big dashboard in front of the
digitally connected. At the same time, the
user. There is only a free flowing form (in
future of working is strongly related to the
the text we will address this free form as a
future of entertainment, because it is both
a ribbon) that flows inside and outside of
done on the same platform. The team be-
the car, creating a loop below the steer-
gan by giving the most general solution for
ing wheel to make a stand for the wheel.
working by providing the user a dashboard
The ribbon flows around the back seat as
which will also be an interactive working
well, providing users a necessary feeling
area. The main display of information will
of a compact cabin like space for EXPE.
be a windshield with augmented reality.
The whole front part of the car is a large

windshield that spreads all the way from
Socializing
the floor of the car, providing a completely
The introduction of autonomous features
new driving experience that was not possi-
in cars brings socialization in the vehicles
ble before autonomous and electrical cars.
to a whole new level. Socialization inside
For the purpose of EXPE presentation,
of a vehicle will be done through technolo-
it was decided to imitate the windshield
gy for remote business meetings, but hav-
display with a screen on which the display
ing a face to face in person conversation
and driving scenario was projected.
is something the team wanted to achieve
Final Documentation
Table 4.6.3.1 Cabin Space mock up
65
66
Final Documentation
Final Documentation
5. Design Specifications
67
68
Final Documentation
In this section we explain our final proto-
functionalities and cable wiring pin outs
type with more details. The section also
were almost exactly the same as discov-
includes the process of making the pro-
ered from the functional prototype. The
totype functional. The subsystems of our
team took the basic knowledge that was
final prototype are:
gained from the functional prototype and

applied it in the same exact way for the
1. Anticipatory Chair
final design, thinking that the chairs would
2. Interactive Steering Wheel
be very similar. The team discovered that

all the pinouts for the motor and encoder
5.1 Anticipatory Chair
lines were the same so the wires were
The chair is comprised of several components: the chair electronics, the firmware
development, and the rotation mecha-
isolated from the rest of the cable harness

and connectors were placed in line to allow easy connection/disconnection from
the main harness.
nism.
Figure 5.1.1.1 displays the system level

view of the electronics portion for the
5.1.1 Chair Electronics

The chair that was used in the final design
was from a 2007 Audi A8L. All of the seat
design and Figure 5.1.1.2 displays the

physical layout of the hardware.
Car Ba6ery
(12V)
Chair Motor
Drivers
Power Supply
(5V, 12V)
Chair
Motors
Microcontroller
RotaAon Motor
Driver
Chair Motor
Sensing
Force Sensor
Board
RotaAon
OpAcal
Encoder
Force Sensors on
Chair
Figure 5.1.1.1 System level view
RotaAon
Motor
Final Documentation
69
Figure 5.1.1.2 Physical layout of the hardware.
5.1.1.1 Microcontroller Board
the reference design to the actual design
The heart and sole of the electronic
was that the team eliminated the use of a
hardware design was the microcontroller
second microcontroller to use as a serial
board. The functional prototype used an
programmer for the atmega2560. Pro-
Arduino Atmega2560 board to control all
gramming was performed via ISP, in-sys-
the motors and sensor inputs. For the final
tem programming, which is standard for all
design, the team decided to move away
Atmel microcontrollers. The main benefit
from the Arduino IDE and use the actual
of designing a custom circuit board is that
Atmel Studios IDE to have more control
the connectors and female socket head-
and flexibility over the microcontroller.
ers to all pins from the microcontroller are
The Arduino wrapper libraries are very
implemented for faster software debug-
handy, but for this final design, the team
ging and modularity since all features and
wanted more control over the software
functionalities of the system were un-
and wanted to eliminate all of the behind
known at the time of designing the boards.
the scene actions that the wrapper library
Figure 5.1.1.1.1 displays the physical
performs. The team used the reference
circuit board for the microcontroller. The
designs of the Arduino atmega2560 as a
schematics can be found in Appendix.
basis to develop their own microcontroller

board that was suitable for their application. One of the major differences from
70
Final Documentation
Figure 5.1.1.1.1 - Physical circuit board for the microcontroller
The microcontroller has three ways of re-
tions for all power/signal lines to all other
ceiving 5V power: a DC barrel jack, a USB
boards and to the chair.
cable, or a power supply. There is a circuit

designed to select between the power
5.1.1.2 Chair Motor Driver Board
sources with the power supply having
The motor driver board for the chairs mo-
top priority if all of the sources are con-
tors is exactly the same as designed for
nected. The board also houses an FTDI
the functional prototype. H-Bridge circuits
RL232 USB to UART component used
were created out of DPDT switches to
for debugging purposes and serial com-
allow for the motors bi-directional abilities.
munication to the laptop that is controlling
Shottky diodes were placed in reversed
the interactive display modes of manual or
biased across the coils and the motors to
autonomous mode on the steering wheel.
help reduce the inductive kickback when
As mentioned above, the board has
the motor was shut down. One major dif-
female header sockets to every pin on the
ference from the functional prototype to
microcontroller to allow for quick software
this final design is that the SPDT switches
development, but has Molex Microclasp
used to connect +12VDC or GND to the
connectors used for modular connec-
supply power to the motor were replaced
Final Documentation
71
with a power transistor. The team noticed
a better experience in the chair. Figure
from user testing that the stop and start
5.1.1.2.1 displays the revised schemat-
motion of the chair movements were very
ics with the power transistor as the power
noticeable and not smooth. The team
enable component and Figure 5.1.1.2.2
decided that having a smooth ramping
shows the physical circuit board that was
function through pulse width modulation
created.
(PWM) was necessary to give the user

Figure 5.1.1.2.1 - Schematics
Figure 5.1.1.2.2 - Physical circuit board
72
Final Documentation
Since transistors were used to supply
5.1.1.3 Rotation Motor Driver Board
current to the motors, heatsinks were also
The rotation motor driver board is essen-
incorporated into the revised design to
tially the same as the chairs motors driver
dissipate the heat generated from cur-
board except that a larger heatsink and
rent flow through the component. Thermal
a dual fan assembly structured orthogo-
calculations were conducted to determine
nally was incorporated. The motor that
the suitable power dissipation rating. The
was used for the rotation mechanism had
maximum current load was measured to
a larger operating current of over 10-15
be around 9 amps, which was the stall
amps. A specific current load was not
current of the motors. The heatsink that
measured since the team did not have the
was used was an Ohmite RA-T2X-25E.
resources to measure large current loads.
Although this heatsink did not meet our
The Ohmite RA-T2X-51E heatsink was
thermal specifications based on power
used in conjunction with a dual fan assem-
dissipation at temperature rise and ther-
bly, as shown in Figure 5.1.1.3.1, because
mal resistance at natural, the team made
the heatsink alone was not able to dissi-
software modifications to be able to detect
pate the heat fast enough if the motor was
stalls within the motors and to immedi-
running continuously. The dual fan assem-
ately disconnect power to the motors. The
bly has one fan blowing from the top of the
heatsink for continuous stall loads of 9
heatsink and one fan from the side. The
amps is definitely insufficient, but for short
fans were operated on the 12VDC sup-
peak durations its suitable. The heatsink
ply line. With the fan assembly, the motor
for continuous loads of around 4 amps,
could run continuously for a much larger
normal operating current load, is fine.
duration than without the additional cooling before it gets noticeably warm.
Figure 5.1.1.3.1: Dual fan assembly for rotation motor heatsink
Final Documentation
73
5.1.1.4 Chair Motor Sensing Board
range that would be compatible for most
The motor sensing board for the chairs
microcontrollers. Figure 5.1.1.4.1 displays
motors is the same schematic design as
the schematic circuit design and Figure
constructed for the functional prototype.
5.1.1.4.2 displays the physical circuit
A comparator was used to level-shift the
board that was created.
encoder output square-wave to a 0-5V

Figure 5.1.1.4.1: Motor position sensing circuit schematic
FIgure 5.1.1.4.2 Motor position sensing board
74
Final Documentation
5.1.1.5 Rotation Optical Encoder
5.1.1.6 Force Sensing Resistors
An optical encoder was used to determine
During the functional prototype, square
the position of the rotation mechanism. A
FSR sensors were used and placed in
one inch 360 degree circle with black and
areas that would be beneficial to obtain
white stripes were used to obtain a square
weight distribution data. Data collected
wave pulse from 0-5V from every color
from the prototype seemed reasonable
transition. The circle of stripes alternated
but only for certain body types and for only
colors with every degree and was as wide
common cases. The team wanted to de-
a encoder tape sensor. This type of optical
sign a system that would be able to detect
encoder was more effective than attach-
and accommodate various body types us-
ing one to the shaft of the motor because
ers had, and change the thresholds based
the position had to be known relative to
on that body type. For example, tall/wide
the turntable that the chair was on and not
users would have a high threshold level to
the motor. If the encoder was connected
meet since it is reasonable to assume that
to the shaft, there were factors that made
more of their body weight would be press-
the source unreliable since the motor shaft
ing the sensors. One major finding from
was connected to a gearbox and then
the functional prototype was that the small
a vertical shaft. The encoded circle was
square FRSs were insufficient in gather-
placed underneath the turntable.
ing data from different body types so a

different kind of FRS was used. Figure
5.1.1.6.1 displays the initial configuration
of FSRs that were tested with for the final
design.
Figure 5.1.1.5.1 Rotation encoder - optical
Figure 5.1.1.6.1: Initial configuration of FSR
sensor and circle of stripes
tested
Final Documentation
75
Long, thin FSRs by Interlink (408 series)
directly on the chair back side cushions.
were used in conjunction with the square
Since these cushions compress and
FSRs to help with accommodating various
expand naturally, it was hard to get an
body types. The 408 series only measure
accurate reading since slight user adjust-
the total amount of force on the sensor
ments would cause unpredictable devia-
and not the location where the pressure is
tions from a calibrated zero reading. For
applied. The longer sensors for the sides
the final design, only the chair seat back
and the shoulders are beneficial for height
side sensors and the two middle square
detection since it had a much larger area
sensors were used for rotation and tilting
to sense shorter and tall people within a
back respectively.
fixed position.
Initially, all the sensors shown in Figure
The circuit design was exactly the same
5.1.1.6.1 were used to differentiate the
as from the functional prototype. The
tilting and rotating functions but after
same board was actually used but with
threshold testing with various users, it
slight modifications. Two extra sensor
was evident that more sensors were not
slots were added to accommodate the
necessarily the best solution for distin-
sensors for the sliding function via the
guishing different intentions. Having less
footpad. Figure 5.1.1.6.2 displays the FSR
sensors was better since it helped reduce
protoboard and Figure 5.1.1.6.3 displays
the amount of misfires from uncontrol-
the circuit schematics.
lable factors. The sensors on the bottom

cushion were not used in the final design
due to many factors that could not be
solved by simple software algorithms. One
observation during the threshold testing
was that users would often have objects in
their pockets, which caused the readings
from the sensors to not accurately reflect
the users body. Another key observation
was that users do not sit symmetrically on
the chair nor do they sit within a consistent pattern. Most users tend to sit more
toward the right side with more pressure
applied on the right side. This could have
been compensated for within software but
another uncontrollable factor that caused
problems with the sensor readings was
that the 408 series FSRs were placed
Figure 5.1.1.6.2 FSR protoboard
76
Final Documentation
CD4067BE
OUT
T2
T1
B6
B5
B4
B3
B2
B1
CONTROL_A
CONTROL_B
GND
OUT
IN7
IN6
IN5
IN4
IN3
IN2
IN1
IN0
A
B
VSS
VDD
IN8
IN9
IN10
IN11
IN12
IN13
IN14
IN15
INHIBIT
C
D
+5VDC
T3
T4
T5
T6
GND
CONTROL_C
CONTROL_D
TI - CD4067BE
Figure 5.1.1.6.3: FSR circuit schematic
5.1.1.7 Override Switches

From the functional prototype, the team
observed that users felt more comfortable
having some kind of override switch so
that if the chair does something that the
user didnt want it to do, the user could
use the switches to stop it and correct it.
The original adjustment switches located

on the side of the chair were the most
reasonable to use as an override point.
Just like from the functional prototype,
the resistance from each throw from the
momentary SPDT was measured as 391
Figure 5.1.1.7.1: Switch override circuit schematic and protoboard
Final Documentation
77
ohms and 822 ohms. A resistance of 470
switch are connected directly as input I/O
ohms in series with the 5VDC power line
pins. The enable/disable system switch is
to the actual switch enabled the voltage
connected to an external interrupt pin that
level for the 391 ohm side as 2.36V (ADC
determines whether the switch is in a high
10 bit value of 466) and a voltage level
or low state. The control box also features
of 3.30V for the 822 ohm side (ADC 10
an LCD screen so that one can determine
Figure 5.1.1.7.2: Switch override protoboard
bit value of 652). The protoboard that
exactly what state the system is in for
was made connects directly to the origi-
debugging and normal operation modes.
nal switch adjustment connector. Figure
Figure 5.1.1.8.1 displays the master con-
5.1.1.7.1 displays the circuit schematic
trol box.
for one of the switch and Figure 5.1.1.7.2

displays the protoboard that was made.
5.1.1.8 Master Control Box
A master control box was designed mainly
for the convenience of having users try out
the system at EXPE. The control box has
several main functions: allows the system
to be reset (software) without having to
touch the reset button on the microcontroller board, to calibrate the FSR thresholds depending on users body type, to
reset the chair to the default position, and
to enable/disable the code functions. The
software reset switch is directly connected
to the reset pin on the microcontroller.
The chair reset and the FSR calibration
Figure 5.1.1.8.1: Master control box
78
5.1.2
Final Documentation
Firmware Development
The basic firmware process flow can be

seen in Figure 5.1.2.1. The code can be
seen in Appendix XX.
FSR

Calibrate
Switch
Override
Disabled
?
Determine
thresholds
Based on
body type
Reset chair to
default pos
Check
override SW
touched
System

Enable
?
Y
Reset

Chair
?
Auto

Mode
?
Drive

Mode
?
Read FSR
Move into
drive pos
SLIDE, TILT,
OR
ROTATE
Figure 5.1.2.1: Firmware process flow
SW

touched
?
Move chair
motors
accordingly
Final Documentation
The code first determines whether the
79
cycles through each switch.
FSR calibration switch or the reset chair

switch from the master control box has
If the switch override is disabled and
been turned on. If the FSR calibration
system functions are active, then the
switch has been turned on, the FSR sen-
retractable steering wheel position dic-
sor values are collected through the ADC
tates whether the user is in autonomous
and the users body type is determined
or drive mode. If the steering wheel is
based on a certain threshold level. The
in drive mode, the chair will proceed
body type is categorized as short, me-
into a drive position setting only with all
dium, tall for the height and thin, aver-
other functions deactivated. If the steer-
age, and wide for the width. The height
ing wheel is in autonomous mode, the
category is used to determine the tilt back
chair reads the FSR sensors and checks
thresholds while the width category is
against three different functions: a tilt
used to determine the rotation thresholds.
back, a slide, and a rotate function.
If the reset chair position switch is turned

on, then the chair moves each of the
5.1.2.1 Tilt Function
motors to its zero position regardless of
The tilt function is activated once the user
current encoder position. Once the motors
holds the two centered back FSRs above
stall, then the chair resets to a default
a certain pressure value for a certain
position setting.
amount of time. This helps eliminates a

lot of misfires due to small adjustments
The code is separated into several
made by the user. If the chair back moves
categories that include switch override,
backwards, the user must lean forward
system enable, auto mode, and drive
slightly to make the FSR values go below
mode. The switch override system is
a certain threshold level to stop the chair.
necessary to enable the user to be able
The same actions also correspond when
to stop the chairs movement if it does
the chair moves forward and the user
something that the user doesnt want it to
wants to stop the chair. While the tilt func-
do. If the switch override is enabled, then
tion is enabled, no other function can be
the software checks whether any of the
activated.
switch adjustments on the chair side has

been active/touched. If no switches were
5.1.2.2 Slide Forward/Back Function
touched, the code cycles back through the
The slide forward/back function is acti-
main loop, but if they were touched, then
vated when the user pushes or pulls the
it determines which switch was activated
footpads just like the tilt function. To move
and moves the corresponding motor as
backward, the user must push forward
long as the switch is active. One switch
twice while holding the second push. This
is checked every 40 ms via a timer and
slight impulse nudge is necessary to help
80
Final Documentation
distinguish between sliding and tilting. Us-
5.1.2.4 Move Motor Function
ers tend to push with their feet when tilting
The code was structured to be able to
backward, and would cause a misfire on
take advantage of a PID motor controller
the sliding function. Having this slight
function. Although PWM gave the team
impulse before holding the second push
speed control over the chairs motor,
helps distinguish the user intentions with-
having a way to determine at what rate
out much effort. The code determines if
the speed should be controlled was an
the slight nudge is conducted and if there
issue. A PID controller helps take into ac-
are X amount of samples above a certain
count how fast the microcontroller should
threshold. If so, then it goes into a waiting
increase or decrease motor speed until
period. If the second push is conducted
it reaches its targeted set position. The
between the waiting period then the user
move motor function first determines the
will slide back, if not, then the user will
delta position from its set and current
have to conduct the nudge again. While
position. Once the error is calculated, the
the slide function is enabled, no other
result of the PID equation can be cal-
function can be activated.
culated and the PWM value will change

according to how far away the current
5.1.2.3 Rotate Function
position is from the set position. Once the
The rotate function is activated when the
motor reaches its targeted value, the mo-
user shifts their body to either the left or
tor would be disabled and the move motor
the right side. When the user is facing

toward the front, the user is restricted to
turn to the right up to 180 degree. When in
rotation mode, the tilting and sliding functions are disabled because of the restrictions with the steering wheel clearance.
If the user initiates rotation mode but the
chair is outside of the steering wheel
clearance boundary, the chair moves to
a set clearance position prior to rotating.
Once the user is rotated, the user would
relieve some of the pressure from the
activating side to stop or rotate all the way
up to 180 degrees. To exit out of rotation
mode, the user must rotate towards the
front all the way to 0 degrees for the tilt
and slide back function to be re-enabled.
Figure 5.1.3.2.1: Chair Dimensions
Final Documentation
for that specific motor would be complete.
81
5.1.3 Rotation mechanism design
This is an effective function since it does

not require the code to be constantly loop-
5.1.3.1 Assumptions:
ing through this function all the time. Other
Chair Weight = 30 kgs
functions and tasks can be interweaved
User Weight = 100 kgs
between move motor functions to allow
Rotation mechanism weight = 20 kgs
better utilization of hardware resources

and time.
5.1.3.2 Design calculations
Figure 5.1.3.2.2: Design Calculations
tric
ome a
e
g
on ees in
ia
sed
s ba 80 degr of inert
i
s
i
h
1
t
T
n
e
.
t
e
r
rota
chai
t, it
mom
ng
f the eeds to ase for er is lyi ir weigh
o
e
n
s
s
c
a
e ba
hair
he u
e ch
orst
of th el. The c l. The w ed and t t and th
r
e
t
n
e
e
h
b
s
e ce
whe
weig
g wh
ke a
55cm
at th steerin teering sition li e user
1
s
=
i
l
s
r
s
o
e
th
ai
3 0 kg
e ch and th lear the a flat p tion of is.
ons
i
h
=
s
t
'
n
r
d
u
n
e
i
m
fo
ib
ar
ll c
ax
ass
(dim
hbo
air is
d sti
tion
distr
ned
kgs
of m
rota the das ition an n the ch inuous e desig
0
)
s
f
3
t
o
1
n
2
he
t th
ont
pos
axis
one
m=
rom
-m^
ass
The raints f eclined axis is w ing a c ng abou
omp
2 kg
c
m
2
.
f
m
0
i
t
m
2
r
o
t
s
=5
b
cons ortable bout the it. Assu ab rota
hani
r sla
(l ')
l
n
f
a
mec
2 )+m
gula
n
n
o
a
com lations clined o 30 kgs s
i
t
t
c
1
e
u
m (b2 + l
rota
a re
calc letely r ed as a
ms
g of
c
ts of
n
i
s
0
t
i
=
p
4
s
s
n
t
i
i
s
on
12
hif
com e imag
l'=
(con
lel s
am c
b
xis is
disc
aral
a
iagr
r
p
d
can
n
a
i
l
+
s
s
mas
circu f the ma
ete
l axi
ped
mpl t
o
nd a
ntra
o
a
m
t
e
c
f
u
)
c
i
l
e
h
s
a
S
n
ut a
The
as o cided th eed.
56cm 35cms .
t
abo
e
=
I
s
e
b
M
s
sp
d
=
and
d wa . It was nstant
lab =
2
us r
e
s
i
^
e
r
d
p
m
o
a
a
c
l
s
c
r
2
5 kg
d/se
at a
tion
and
angu
' l =1.22
rota f 0.6 ra move
rect
m
e
m
h
o
d
u
t
l
f
d
im
^2
d of
shou
MI o
spee
c= 2
g-m
max
spee
red ngular and the
r dis
45 k
i
a
4
s
m
l
.
e
u
u
1
a
d
c 2
5
circ
um
axim
the
1 se
tia =
f the
s
hair maxim eed in
he m
/
iner
c
t
l
f
d
d
e
o
a
a
MI o
p
n
an
r
nt
to
rs
wiv
-m a
m at
ome
he s comes angula d be 0.6
N
t
N
m
7
n
l
o
t
l
m
.86
0.75
Tota
one
wou
imu
. Tha
= 30
ue is lues we
6
hen
sts d econds his max
q
.
t
r
e
0
t
o
n
t
x
va
d
ve t
d on
11 s
ratio
uire n those
.445
q
1
e
5
Base ution in ld achie r accele
r
o
is
l
u
la
t the
fety
hen
revo hair sho
angu
s tha tor of sa
air t
n
h
m
c
a
c
u
e
e
c
the
axim
rn th
ich m nal fa
he m
to tu 6 rpm.
1 wh additio
:
d
0
So t
e
4
ir
f
n
y
atel
requ
ion o
ing a
duct d apply 0 rpm.
rque pproxim
e
o
r
t
a
0
The ion is a
t an
at 3
has
t
ear ing poin 1 N-m
g
rota
m
t
r
is
ra
oint
d wo
ope
lecte That is rating p
e
s
The f 240. he ope
o
t
rpm ed that
d
i
c
e
d
82
Final Documentation
5.1.3.3 Initial rotation prototype

The first prototype of the rotation mechanism was designed using a 40:1 reduction
worm and a worm gear from McMaster
Carr. The connecting components were
machined.
A motor was selected from AmpFlow.
This motor operates on 24V and has a
stall torque of 5 N-m and a no load rpm
of 5600 at 24V. These specifications far
exceed the calculated ones including a
factor of safety. The team planned to run
Figure5.1.3.3.1 - Rotation first prototype 1
Figure 5.1.3.3.2 - Rotation first prototype 2
Final Documentation
83
this motor at 12V. This made it easier to

use transistors for PWM switching of the
motor.
There was a lot of noise and alignment
issues with this mechanism. This also
caused a lot of vibration which could be
felt by the users as it was being transmitted through the platform and onto the
chair that was connected to it.
Figure 5.1.3.4.1 - Rotation mechanism initial

assembly
5.1.3.4 Final Rotation Mechanism

It was decided to go for an industrial worm
gearbox in order to remove the noise,
vibrations and resolve all alignment issues
related to the gearbox. Flexible couplings
were used at the input and output shafts
of the gearbox to connect the motor to the
gearbox and to connect the gearbox to the
platform. Figure 5.1.3.4.1 shows the first
assembly of the gearbox.
Figure 5.1.3.5.3 - Rotation connection to motor
Figure 5.1.3.4.2 - Rotation connection to platform
84
Final Documentation
5.1.4 Cabin Base Design

The cabin base was constructed in three
parts for ease of transportation. The front
part was used to mount the dashboard,
the middle part housed the rotation mechanism and the chair and the third part was
used to mount the back seat. The first and
third parts were composed of a normal
rectangular support design. The middle
part was designed in a different way to
transmit the loads from the rotation more
uniformly. The turntable was supported at
eight points on an octagon support right at
the center of the base. Appropriate spacers were used to align the turntable and
make it flat with respect to the ground.
Figure 5.1.4.1 - Cabin base
Figure 5.1.4.2 - Drawing of the central cabin base design
Final Documentation
85
5.1.5 Footpad Design

The footpad was constructed in three layers (each a 0.25 inch in thickness) - the
base layer, the middle layer that serves as
a support and an upper section consisting
of the actual actuation element/section
and the housing around it. The footpad is
22 long and 19 wide. The three layers
have been shown in the following figures 5.1.5.1; 5.1.5.2; 5.1.5.3
The middle layer has a slot in it to route

the wires of the two force sensors that
have been placed at the front and back of
the footpad.
Figure 5.1.5.2 - bottom footpad layer
Figure 5.1.5.4 - Footpad top view showing wire routing
Figure 5.1.5.1 - top footpad layer
Figure 5.1.5.3 - middle footpad layer
86
Final Documentation
Direct actuation of the force sensors using
altered to also consist of rounded protru-
the edge of the actuating section does not
sions to make it easier to slide while also
work well, since it is pretty much acrylic to
maintaining alignment. To further reduce
acrylic contact with the force sensor be-
the sliding friction, nine ball bearings were
ing in between. The force readings were
added right below the actuating section,
much more reliable when a rubber stop-
embedded in the middle layer. To make
per was placed in between the actuating
it less slippery and much easier to actu-
section and the force sensor. As seen in
ate the footpad, the top was covered with
the diagram for the top layer, the front
vinyl rubber and grip tape to increase the
and back of the actuating section were
friction between the users feet and the
also changed to rounded protrusions for
actuating section.
a point contact on the force sensor. This

made the force response also much bet-
In terms of moving across the platform,
ter. In the next iteration, to further improve
the first iteration of the footpad had ball
on the force response of the footpad and
transfer wheels embedded in the bottom
to reduce sliding friction with the sides
layer so it could slide across the platform
of the actuating section, the sides were
on four points of contact.
Figure 5.1.5.5 - Footpad with embedded ball casters
Final Documentation
87
When the carpet was added this sliding
cal push and pull triggers for mode trigger-
led to a lot of friction, thereby increas-
ing. The steering wheel is a new design
ing the torque required by the rotation
for a sleeker aesthetic and is mounted to
mechanism to turn the chair. The final
this retraction mechanism.
iteration of the footpad does not consist of

the caster wheels and is supported by a
Steering Wheel
structure connected to the turning platform
Parts for this Interactive Steering Wheel
that is rigid enough to not let the footpad
were created with a 3D printer (Picture
touch the ground when the users feet
bellow), except for the bottom part which
are resting on it. This makes the footpad
was manufactured with a CNC mill. The
essentially float above the platform while
interactive screen was made using a used
rotating, which leads to a much better
Android tablet.
torque efficiency.
5.2 The Interactive Steering Wheel
Parts in this mechanism are hand made
This subsystem is a combination of the
using various manufacturing processes
retraction mechanism and the steering
including lathe, welding, sheet metal cut-
wheel. The retraction mechanism is a con-
ting and bending.
struction that allows the user to do physi-
Figure 5.2.1 - 3d Printing of Steering Wheel
88
Final Documentation
5.2.1 Structure of The Interative

Steering Wheel
Interactive Steering Wheel
Figure 5.2.1.1 - Steering Wheel
Handles
Frame
Android Tablet
Figure 5.2.1.2 - Steering Wheel Parts
Final Documentation
89
Back Lid
Figure 5.2.1.3 - Steering Wheel Parts Back
Ejection Spring
Tightening Band
Switch
Outter shaft
Inner Shaft
Base
Microswitch Support
Lid
Figure 5.2.1.4 - Retraction Mechanism Parts
Flange
90
Final Documentation
5.2.2 The Making of The Interactive
ties with a modification friendly operating
Steering Wheel
system based in linux. The tablet was
The CAD model was created in Rhinocer-
rooted, which means installing over the
ous modeling software and divided into
manufacturers firmware with a custom
two categories:
ROM in order to free up more functionali-
Milled parts
3D-printed parts
ties that were not allowed with the manufacturers settings and also sets free the
superuser(SU) functionality in order to
Milled Parts
receive commands over WiFi. In this case,
The Back Lid was manufactured by CNC
the team used a ROM called Cyanogen-
milling machine out of blue prototype
mod 9.1.0 since it was easily accessible
plastic.
and the instructions were available.
3D-Printed Parts
Finishing
The maximum size of Aalto Digital Design
After printing the parts, they were glued
Laboratorys uPrint printing envelope was
together and finished. The parts were
150x210 mm, so also the model had to be
filled and sanded before they were
splitted into several pieces, which together
painted. After filling, sanding and painting,
formed the Frame and Grip parts. The
the tablet was inserted into the back lid
model was generated into 14 parts that
and screwed on the frame.
were connected with studs that fit perfectly

to the corresponding holes. The settings
of the prints was to print solid ABS and
Base
altogether 561,39 cm^3 build material
This part of structure was CNC milled
and 197,43 cm^3 support material was
from blue prototype plastic and aluminum.
used within 44,47 hours of printing. After
The Base is made out of plastic and the
printing the parts, they were washed in an
Tightening Bands are made out of alu-
ultrasound bath in order to dissolve the
minum. The 3D models were made with
support material from the actual ABS.
Creo CAD software and those models

were analyzed with Mastercam which
Tablet
made the milling paths numerically for the
The tablet was selected by the size, the
CNC milling machine. After milling, two
operating system and the price. Gal-
M10 pressing screws were put on both
axy Tab 2 was one of the most obvious
tightening bands. The tightening bands
selections since it has 10.1 large screen,
are attached to the base with M5 bolts.
android as a operating system and is
Figure n shows the manufactured parts
one of the cheapest in its price category.
and the dimensional drawings can be
The android is well known for its capabili-
seen bellow.
Final Documentation
91
Figure 5.2.2.1: Base
Figure 5.2.2.2: Base DImensions
92
Figure 5.2.2.3: Tightening Bands
Figure 5.2.2.4: Tightening Bands Dimensions
Final Documentation
Final Documentation
93
Shaft
The shaft was manufactured with a lathe.
Here are the steps for making the shaft
part of the Retraction Mechanism.
Figure 5.2.2.5: Shaft
1. Step
Steel blanks were cut into correct lengths
2. Step
A hole for the guide nut was drilled on the
inner shaft, threads for locking bolt were
made and the guide nut was glued into its
spot.
3. Step
Grooves were made on the inner shaft
with a lathe.
4. Step
The flange was welded onto the inner
shaft and four M5 holes were drilled into
the flange.
5. Step
Guiding profile were manufactured and
M10 threads were welded onto Outer
Shaft.
6. Step
Locking screw was with a M6 bolt, two M6
nuts, two steel washers and one plastic
washer. The spring is also attached to it.
94
Final Documentation
Figure 5.2.2.6 - Inner shaft technical drawing
Microswitch Support
Cover for the retraction shaft
This support was manufactured with a
The cover was made of an ABS plastic
sheet metal cutter. The support is de-
tube, which had a hole cut in order to
signed to house the microswitch XGG2-
slide over the shaft. At the end of this tube
88-S20Z1 ,which can be seen in Figure n.
there was a glued laser cut flange to cover
Figure 5.2.2.7 - Microswitch support
Final Documentation
95
where the retraction mechanism met the
pictures of different recognizable appli-
dashboard. On top of this ABS plastic
cations like facebook, google calendar
covering there was leather covering velcro
or a movie. These applications could be
onto it.
viewed by simply swiping through the options on the tablet.
Assembled Retraction Mechanism

After all the parts are manufactured the
5.2.4. Communications between devices
retraction mechanism was assembled as
Protocols used:
seen in Figure 5.2.2.8
Android Debug Bridge (ADB), wifi port

5555
TUIO protocol, wifi port 3333
http, wifi port 80
Universal Asyncronous Receiver Transmitter, UART, Physical USB cable
The setup can be seen in the schematics
picture below. The trigger for the mode
Figure 5.2.2.8 - Assembled Retr. Mechanism
5.2.3 The interaction specifications of

the steering wheel
General overview of the setup and functionalities
The interaction part of the steering wheel
was meant to fulfill a simple purpose: to
demonstrate that the windshield can be
utilized as display while the car is driven
in autonomous mode. In order to change
these modes the steering wheel retraction
gave the signal for the tablet to change
the app and indicate the mode change in
words on the tablet screen. The steering
wheel also provide some haptic feedback
by vibrating twhen a mode change occurred. For the windshield, this meant that
there was an additional interactive window
over the driving scenario to represent our
vision of the future of work. There were
change was a physical switch behind the

steering wheel at the end of the short
steering shaft. The switch sent information
to the microcontroller through wires that
the end of shaft when moved to the other
position. The microcontroller used UART
communications to send either the letter
M for manual mode or the letter A for
autonomous mode to the serial port. On
the Mac side, there was a python script
listening to the serial port and determining
what actions to perform. The python script
operated web browsers on the projector through an applescript code and sent
the commands via wifi to the tablet which
was listening to the ADB commands. The
tablet sent back TUIO protocol data for
the computer in order to control the app
on the windshield display by swiping. A
program called Tongseng was used to listen and pass the TUIO data to the caress
96
Final Documentation
webserver. Javascript was used to make
Projector canvas
the webpages functional and for receiving

the touch events sent by Tongseng.
App
Microcontroller
Mechanical switch
behind the steering
wheel
VGA cable
Projector
Mac computer
Caress server
node.js
python script
Tongseng
USB cable
WiFi router
Galaxy Tab 2 tablet

Apps and hardware commands
Data to tablet
Data to computer
Touch events TUIO protocol
Figure 5.2.4.1: Communication schematics
The physical implementation of the setup

can be seen in the following picture.
Figure 5.2.4.2: Physical Implementation
Commands over ADB
Final Documentation
97
Figure 5.3.1.1 - Gluing the pieces
5.3 Open and Clear

Cabin Space
5.3.1 The Making of Open and Clear
Cabin Space
The dashboard:
Laser cutting the Duron pieces:
The team opted to make the dashboard
economically. The shape of dashboard
was first designed in Rhino. It was sliced
into 250 pieces so that laser cut duron
pieces could then be stacked to create the
shape.
Material needed:
3 ft x 4ft 1/4in Duron sheets
8 sheets
Gluing the pieces:
Each piece was numbered for easier assembly. On each piece, there was also a
rastered line to guide where the next piece
should be glued.
Material needed:
Wood glue
Clamps
98
Final Documentation
Sanding and applying filler
Coloring
The team next spent 2 weeks sanding off
The dashboard was finished by applying
the edges and applying Bondo to make
primer and flat black spray paint. (picture
the surface smooth and curved. (kuva
of dashboard)
taalla) In order to prevent the wood part
Material needed:
from cracking when force was applied, the
6 x primer
dashboard was then covered in fiberglass.
6 x flat black spray paint
Material needed:
Sanding machine
Grinder
40 pieces of 60 sand paper
20 pieces of 150 sand paper
3 cans x wood filler
5 cans x Bondo
Fiberglass kit
Figure 5.3.1.2 - Sanding the dashboard
Final Documentation
99
Figure 5.3.1.3 - Painted dashboard
The Ribbon
Making the bends
The ribbon was divided into 4 pieces - two
The team experimented with how to best
in the front and two in the back. The initial
make a uniform curve that could be rep-
ribbon thickness was supposed to be
licated easily. One of these ways was to
approximately 3cm. However, making the
heat up foam and to bend it over a mold.
ribbon this thick was quite a large chal-
However when the foam was heated
lenge and so alternatively materials and
instead of becoming soft, it shrank and
dimensions were explored.
became hard.
Another solution was to divide the rib-
Laser cutting acrylics
bon into smaller sections and heat each
The .dxf files were sent to a water jet cut-
acrylic piece in an oven or with a heat gun
ting company to have two 1cm thick black
to bend it. It was impossible to bend 3cm
acrylics sheets cut. (picture of sheets)
thick acrylic. The team decided to use
2 - 4 x 8 black acrylic sheets.
8mm thick acrylic pieces, which could be

bent with a heatgun.
100
Final Documentation
Figure 5.3.1.4 - Banding the ribbon
Because of the restrictions in time and the

amount of work, the team chose to use
thinner acrylic that could be heated and
bent. Since the pieces were rather large, a
grill was used to heat the entire section of
the acrylic and was molded over a trashcan to proper radius.
Painting
The ribbon was spraypainted flat black to
match the dashboard.
Material needed:
ports. The wood pillars were inserted into
1 piece of 600 sand paper
the plastic pipes, which were mounted and
2 x primer
screwed to the platform.
2 x flat black spray paint
A piece of 2x4 wood was attached to the

bottom of the ribbon. It made gluing the
Mounting the ribbon
wood to the plastic easier, as well as at-
The ribbon is created to stand on pillars to
taching the pillars to the plastic. These pil-
keep it at the right height. Wooden pillars
lars got screwed to the into each wooden
and plastic pipes were used as the sup-
board.
Figure 5.3.1.5 - Barbecuing the ribbon
Final Documentation
101
To attach the ribbon to the floor, the team

drilled holes in the floor and then placed
small plastic pipes through those holes.
Below the floor, we glued those pipes to
a small wooden board that we screwed to
the floor from below
Ribbon
2x4 Wooden Board
Wooden Pillars
Plastic pipe
Figure 5.3.1.6 - Mounting the ribbon
102
Final Documentation
Final Documentation
6. Project Planning and

Management
103
104
Final Documentation
6.1 Deliverables and Milestones

Date
Deliverables
10/22/12
Corporate Project Team Formation
11/8/12
Benchmarking review
11/29/12
Critical Function Prototype
12/4/12
Fall project brochure
12/6/12
Fall presentation
12/11/12
Fall documentation due
1/23/13
First meeting with Audi contact and engineers in Ingolstadt
1/24/13
Second meeting with Audi HMI lab representative in Ingolstadt
1/18/13 - 1/29/13
Dark horse prototyping
1/31/13 - 2/14/13
Funky prototyping
2/15/13 - 2/28/13
Functional prototyping
3/3/13
First vision locked
3/10/13
Turning point brochure
3/12/13
Turning point presentation
3/13/13 - 3/16/13
Aalto in Geneva Motor Show
3/21/13
Winter documentation due
3/22/13 - 3/31/13
Stanford in Helsinki
3/27/13
First whole team meeting with Audi locking concept
4/18/13
X is finished
5/12/13
Aalto in Stanford
5/16/13
Penultimate is due
5/28/13
EXPE brochure is due
6/6/13
EXPE
6/11/13
Final documentation due
Table 6.1.1 - Deliverables and milestones
Final Documentation
105
6.2 Budget and

spendings
This section includes the money spent for
the project so far, for both Stanford and
Aalto. The funds remaining from fall term
for both teams will rollover into the winter
and spring quarter budgets.
Date
Vendor
Item
Purpose
Amount
11/1/2012
Amazon
Mindflex Game
Benchmarking
$55.99
Stanford
Batteries and
Materials for
Bookstore
markers
Benchmarking
Sparkfun
Accelerometer
CFP
$38.23
Item
Purpose
Amount
Van renting
CEP
38.95
Saitek Aviator PC
Steering Testing
49.90
Steering Testing
24.90
Steering Testing
57.10
11/4/2012
11/21/2012
$30.44
Table 6.2.1: Stanford Spendings - Fall
Date
14/11/2012
Vendor
Mechanical
Engineering Guild
27/11/2012
Verkkokauppa
27/11/2012
Verkkokauppa
Gran Turismo 5
PS3
Thrustmaster
27/11/12
Verkkokauppa
Experience
Racing Wheel
28/11/12
ABC Gas
Diesel gasoline
Audi A6 Test
Station
for Audi A6
Drive
Table 6.2.2: Aalto Spendings - Fall
26.93
106
Final Documentation
Date
Vendor
Item
13/2/2013
ABC gas
Gasoline
27/2/2013
Hobbyfactory Oy
Servo
27/2/2013
Protoshop
Aluminium pipe
27/2/2013
Motonet Oy
Seatbelt
27/2/2013
Motonet Oy
Purpose
Funky Prototype
Testing
Functional
Prototype
Functional
Prototype
Functional
Prototype
Amount
31.07
71,9
18.23
39,9
Cover for the
Functional
steering wheel
Prototype
Item
Purpose
Amount
Paper Bots
26.86
Paper Bots
97.75
Dark Horse
180
Dark Horse
48.87
Dark Horse
Dark Horse
15
Travel
27.7
7,9
Table 6.2.3: Aalto Spendings -Winter
Date
Vendor
10/1/2013
JAMECO
1/23/2013
Frys
1/25/2013
UHAUL
1/25/2013
Frys
Dark Horse
1/26/2013
Testing
Cash
Dark Horse
1/26/2013
Testing
Cash
1/25/2013
Zipcar
Table 6.2.4: Stanford Spendings -Winter
Electronic
Components
Electronic
Components
Renting the van
Electronic
Components
Participant
incentive
Participant
incentive
Travel
Final Documentation
107
Date
Vendor
Item
Purpose
Amount
1/25/2013
UHAUL
Parking
Dark Horse
5.1
1/28/2013
UHAUL
Gas
Dark Horse
33.14
Electronic Components
FUNKY
62.89
2/20/2013
DigiKey Electronic
Components
2/20/2013
Jameco
FUNKY
47.52
2/10/2013
Frys Components
FUNKY
4.96
2/10/2013
Radioshack
FUNKY
35.21
2/12/2013
McMaster Carr
Hardware
FUNKY
50.31
2/8/2013
AUDI Gas
Gas
AUDI testing
79.72
Fabric
FUNKY
11.3
2/10/2013
JOANN Fabric and

Craft Stores
2/9/2013
FRYS
Functional
15.14
2/13/2013
SparkFun
Functional
50.74
2/20/2013
Andy Mark
Functional
130.18
Resistors
Functional
3.24
Switches
Functional
13.86
Functional
33.04
Functional
13.02
3/11/2013
3/3/2013
Radioshack
Resistors
Radioshack
Switches
DC motors and
connectors
Fasteners and Shaft
2/25/2013
McMaster Carr
2/27/2013
Ace Hardware
2/25/2013
JAMECO
Functional
44.57
2/26/2013
Ace Hardware
Hinges
Functional
29.25
2/25/2013
Radioshack
Functional
16.24
Table 6.2.4: Stanford Spendings -Winter
Couplings
Fasteners and Lead
Screw
Electronic
Components
108
Final Documentation
Date
Vendor
Item
Purpose
Amount ($)
2/22/2013
Frys
Functional
40.03
2/22/2013
TAP Plastics
Plastic sheets
Functional
108.85
2/22/2013
ACE Hardware
Fasteners
Functional
33.48
2/22/2013
Radioshack
Functional
15.15
2/13/2013
Staples
Easy button
Functional
7.58
2/20/2013
2/10/2013
Microsoft
Store
Olive Garden
Kinect
Dinner
Table 6.2.4 - Stanford Spendings -Winter
Gesture
Recognition
Food for
SUDS
150
390
Final Documentation
109
Table 6.2.5 - AaltoSpendings - Spring
Balance (EUROS)
Type/Name
5833.33 Bearings
5786.33 Teflon pipe
5760.29 Force sensors
5760.29 Stainless steel pipes
5719.99 Rubber matt
5715.53 Aluminum blank
5671.94 Galaxy tab 2
5222.04 Steering wheel for PC app
5218.05 Stainless steel pipes
5213.96 Spring presser M10, small ball head
5168.21 Spring presser M10, big ball head
5129.65 3D printing SW
4716.52 ACE hardware
4611.39 1/4 '' Duron
4605.22 gearbox
4231.22 home depot
4207.97 tshirts
3924.5 carpet
3403.78 Jameco electronics
3259.1 home depot
3251.1 ACE hardware
3228.81 home depot
3050.41 projector
2935.15 Ribbon pieces
1924.99 miscellaneous product
1915.8 Home depot
1907.25 Home depot
1899.23 Home depot
1833.87 Home depot
1723.96 coals
1692.9 Home depot
1666.63 Home depot
1638.55 potography rental
1587.29 Repeater cable
1561.74 Poster
1461.33 Ace hardware
1442.1 Fry's
1416.99 Ace hardware
1386.88 Jo-Ann Fabric and craft
Total
Date
8.4.2013
10.4.2013
Amount (EUROS)
47
26.04
5.4.2013
17.4.2013
17.4.2013
27.4.2013
3.5.2013
25.4.2013
26.4.2013
3.5.2013
12.5.2013
13.5.2013
16.5.2013
14.5.2013
17.5.2013
18.5.2013
20.5.2013
21.5.2013
22.5.2013
22.5.2013
24.5.2013
25.5.2013
29.5.2013
30.5.2013
30.5.2013
31.5.2013
30.5.2013
30.5.2013
26.5.2013
29.5.2013
29.5.2013
30.5.2013
2.6.2013
2.6.2013
1.6.2013
28.5.2013
6.5.13
1.6.13
27.5.13
4.6.13
5.6.13
5.6.13
19.5.13
40.3
4.46
43.59
449.9
3.99
4.09
45.75
38.56
413.13
63.19
41.94
6.17
374
23.25
283.47
476.72
44
144.68
8
22.29
178.4
115.26
1010.16
9.19
8.55
8.02
15.37
20.9
17.4
11.69
95.41
14.5
31.06
26.27
28.08
51.26
25.55
100.41
19.23
25.11
30.11
19.98
4466.43
110
Final Documentation
Table 6.2.6 - Stanford Spendings - Spring
BALANCE ($) Type/Name

5560.56 Turntable
5284.76 FSR Strip
5263.3 Gas
5178.19 Metal
5066.41 Electronics
4560.01 Motor
4451.7 McMaster
4298.9 McMaster
4270.87 McMaster
4244.25 Adafruit
4231.43 Fry's
4202.22 Harold's Upholstery
3743.94 Alan Steel
3712.61 Ace Hardware
3686.32 Alan Steel
3662.34 GAS
3620.99 Coconuts
3477.06 Adafruit
3464.24 Jameco
3444.29 HomeDepot
3396.6 Gas
3327.73 HomeDepot
3219.1 HomeDepot
3148.28 Rajjot
2894.28 TAP Plastic
2871 JAMECO
2831.89 Digikey
2796.09 JAMECO
2694.7 Digikey
2631.48 JAMECO
2594.76 DIGIkey
2565.4 Fry's
2437.81 TAP Plastic
2193.32 HomeDepot
2163.32 RM 36
1888.32 McMaster
1812.12 Automation Direct
1757.83 HomeDepot
1720.45 home depot
1655.07 ikea
1373.36 home depot
Total
Date
Amount (USD)
04/05/2013
275.8
04/04/2013
21.46
04/02/2013
85.11
04/09/2013
111.78
04/10/2013
506.4
04/10/2013
108.31
04/09/2013
152.8
04/09/2013
28.03
04/12/2013
26.62
4/15/2013
12.82
4/16/2013
29.21
4/16/2013
458.28
4/23/2013
31.33
4/23/2013
26.29
4/23/2013
23.98
4/23/2013
41.35
04/05/2013
143.93
4/16/2013
12.82
4/24/2013
19.95
4/30/2013
27.58
4/19/2013
20.11
4/27/2013
40.55
4/25/2013
22.14
4/19/2013
6.18
4/26/2013
108.63
4/26/2013
70.82
05/01/2013
254
05/02/2013
69.28
05/02/2013
23.28
05/02/2013
39.11
05/06/2013
35.8
05/07/2013
101.39
5/14/2013
63.22
5/14/2013
36.72
05/08/2013
29.36
05/11/2013
127.59
5/13/2013
22.53
05/09/2013
221.96
5/17/2013
30
5/17/2013
275
76.2
19.56
34.73
37.38
65.38
281.71
4256.48
Final Documentation
111
6.3 Distributed Team

Management
Bearing with 10-hour time difference and
different timetable, the team decided to work
on different CFPs. The Aalto team tried to
create a way for easy manual steering, and
Stanford focused on the smooth steering
transition between manual and autonomous
driving modes. The two teams meet once
or twice a week for 30 - 90 minutes on
Google+ hangout to keep each other posted
on current progress, to share ideas, and to
show prototypes. (Table 6.3.1)
Tools
Purpose
Emails
Contact liaison, schedule update, important file sharing
Flowdock
Picture updating, file sharing, showing prototypes
Google Docs
Documentation, data recording
Google+ Hangout
Facebook VW-Audi Group
Discussion, brainstorming, job division, idea sharing,

weekly update
Finnish team internal group for scheduling, quick questions,
idea sharing
Photos
Recording testing process, capturing inspiration
Videos
Documenting tests, conversations, and prototyping
WhatsApp Group
Dropbox
Table 6.3.1 - Comunication Tools
Finnish internal group for instant chat on schedule

confirmation and daily accident report
Stanford internal group for file sharing
112
Final Documentation
6.4 Stanford EXPE

Stanford EXPE is the closing show of
the ME310 course. At this event, each
team has to give a final presentation at
Stanford University and demonstrate the
final product to the crowd in a small sized
booth. Audi Evolve showcased their final
product for the first time through a 12-min
presentation and had 3-hour interaction
with the EXPE visitors. We have made
two videos for EXPE. One was used in
the formal presentation, in which a future
Audi driver is using our product, switching

between two modes and interacting with a
backseat passenger. This video shows all
critical functions of the interactive steering
wheel and the anticipatory chair. The other
video is a detailed product demo. Each team
member gives a short explanation about our
product vision and value. Our Expe booths
featured a type of users manual to explain
how to use the chair and steering wheel, as
well as a tv display for our product video.
Final Documentation
6.5 Future Work
The team has identified future opportunities that this final system has the potential
to include if future iterations were to be
made.
Force Sensor Calibrated Driver Presets
This chair calibrates using the force sensors to determine what body type you
most likely are from your default seated
force readings. Based on these, the chair
could determine who exactly is sitting in
the chair and adjust the driver presets
accordingly. That way the need to input
if you are driver number 1, 2 or 3 like in
todays car settings is eliminated and the
chair simply knows based on you sitting in
the chair.
Similarly if the driver is calibrated as a
shorter person, then the chair can lower
so that the user will be able to have their
feet reach the floor and pedals without
having to do any manual adjustments.
This will make it much easier for shorter
people be able to utilize the foot pad for
moving their chair forward and backward.
This can be applied to taller people as
well who may need the chair raised up
and moved backward in order to reach the
wheel and foot pad comfortably.
These presets can be extended passed
just chair adjustments to mirror adjustment, temperature control, lighting control,
and even media/ radio preferences. This
will ultimately make the car know who is
sitting in the car without the user ever hav-
113
ing to indicate who they are. The car will

become smart and be able to anticipate
the drivers every need.
Reconfigurable Chair Locations
The team sees a potential for getting that
fourth chair back into the cabin space.
It is possible to put all of the chairs on a
railing system so that the chairs can move
around. If there are only two people in
the car then the drivers seat can move to
the left and the back right seat can move
forward. That way the two passengers
are seated next to each other. Similarly if
there are four passengers than the right
back seat can move forward, and another
seat be unfolded from the cabin space
to replace the seat that was just moved
forward. We think that this will allow for
more mobility and flexibility for users in the
cases were this three seat configuration
may not be optimal.
114
Final Documentation
6.6 Personal Reflections
Winter Reflection
Sangram
went through some great prototypes which
Fall Reflection
I think registering for the ME310 sequence
before I graduate from Stanford was the
best decision throughout my stay here. It is
now that I know what I would have missed
if I had not taken 310 this year. Its been a
great learning experience for me and I am
enjoying working with my team at Stanford
and the global team as well. It is still amazing
how we managed to get so much done in
just over a month since serious work on the
project started. There were times when we
were completely lost and there were times
when we knew exactly what we wanted
to do. The former was required but was
completely outside my comfort zone. That is
one important thing I have learnt this quarter
- how to deal with total ambiguity.
This was an interesting quarter for me. We

gave us good insights in the design space. It
took long hours of work to get our prototypes
ready in time and functioning properly, but
our team effort was very good and we got
things done! The best part is that I am
enjoying working on this project so much that
I did not even realize how this quarter went
by. It was when I started considering other
things as being interruptions to 310 work
that I realized how important this project has
become to me. We have come up with a
very good direction for our project. I still keep
wondering if there is any other good idea
out there, which is why I wish we had done
a couple of more dark horse prototypes. I
also think that our communication with the
global team could have been better and we
as a team would have discussed not only
Everyone can manage to think about
the idea for the next prototype but how each
something to do for testing and benchmarking,
of the teams is planning to implement and
but I think the most important part and the
test that idea. In any case, I am excited for
one at which my team is very good at is
the Spring quarter and slightly scared after
sitting down and analyzing the observations
looking at how quickly the next deadlines
and getting tangible conclusions from the
are coming up. But I am very sure that we
tests that were done. I have found these
can get there. I am looking forward to a good
intense what-is-going-on discussions to
final brainstorm with our team in Finland
be the most useful for me as well as for
during Spring break and hope that we come
the team. Observing and thinking about
with a solution directly based off all the good
everyday situations and drawing parallels
insights we have over the last two quarters.
into the technology and situations that

will exist in the future is something which
Spring Relfection
will prove to be very useful while heading
I have a lot of mixed emotions when I
towards a final solution in spring. I think this
write this reflection. ME 310 has been
year team Audi has a great opportunity to
everything that I had hoped for and more. It
create and design for a unique experience
is remarkable what we have achieved as a
for futuristic cars and I am looking forward
team. The final product we have delivered
to getting back to work in January.
Final Documentation
115
speaks for itself but the journey is something
teammates who have become such good
which does not come across and is not so
friends that it is not easy to say goodbye to
apparent. It has been a tireless effort for
them as they leave one by one. It is time to
the whole year. The team went through ups
say goodbye to Stanford. It has been one
and downs and we survived it to probably
awesome ride and the AUDI project has
become one of the best functioning teams
been the best experience so far, learnings
in 310 this year. I think our Finland trip was
and moments that I will cherish for the rest
the game changer. We realized that more
of my life. It is not the end of 310, this is
than half of our issues were communication
just a new beginning for 7 people who I am
issues and we went from scattered thoughts
sure will carry forward this experience and
here and there to a strong unified vision and
achieve great things in life.
a solid plan as we got into the spring quarter.

I enjoyed the design and exploration phase
Stephanie
as it was something new to me and I really
Fall Reflection
learnt a lot of things in the first two quarters.
As fall quarter comes to a close, it is clear
Not to say that I did not learn anything in the

last quarter, I learnt a lot but I was totally
in my comfort zone in Spring. We have
ultimately implemented everything at a very
professional level and have finished what
we set out for. I wish we had the planned
our budget more properly so that we could
have directly invested at the beginning in
that being in a state ambiguity is a fact of

life in ME 310. Our team began extremely
enthusiastic about the prompt and utterly
overwhelmed by the amount of things we
knew nothing about. It has been difficult
thinking about designing more than 20 years
in the future. Deciding on what to assume
and how to even come to those conclusions
about the future has been a long process
I am enjoying
working on this project
so much that I did not
even realize how this
quarter went by.
involving long discussions and debates
the changes that we made towards the end.
communicating with the global team has
But it was all a learning experience and I am

glad I went through it.
That being said, it is now time to
say goodbye..to the loft, to our space, to my
within our Stanford team and with our global

team. I hope we will get the chance to do
more wild brainstorming next quarter, since
this quarter the assignments often times
seemed to hamper our ability to really get
creative.
Furthermore,
collaborating
and
presented many challenges. Scheduling

meeting times and working around 7
peoples schedules has been difficult,
though I think we have gotten better at it.
Also the collaboration problem, I believe has
116
Final Documentation
in large part been because of the difference
tasks so that things just get forgotten on our
in assignment due dates. For a large portion
to do lists. Generally, this quarter has been
of this quarter, the global teams assignments
a lot of fun and I cant wait to see how our
were due about a week after ours. This
project evolves over the next quarter.
made it difficult to brainstorm ideas together

and really talk to each other and learn
Winter Reflection
from each others assignments, in order to
There are many emotions that come to mind
better inform each teams understanding
when reflecting on this quarter: excitement,
of the design space. When both teams
frustration, confusion, clarity, and, well,
were together, the brainstorming sessions
exhaustion. We began unsure of what we
were fun, innovative, and inspiring. I really
were doing and what direction we were
appreciated everyones creative ideas and
headed, but after diving into Dark Horse
hope we will come together more often in
everything seemed to begin to make sense.
the coming months.
I was excited about what we were learning
Prototyping and making an experience

come to life is, for me, the most enjoyable
part of the process. I like being able to
touch and see an idea that was born from
our brainstorming, and I enjoy seeing how
other people react and respond to that
...this has been the

best course I have
taken so far
and the experiences we were creating and

testing. The time has gone by so quickly
and we never let up to catch our breath, but
I think that is what has made this quarter
so interesting. We were always working
our way through a maze of problems or
decisions that needed to be addressed.
The difficulties were in figuring out how to
communicate everything we were learning
with our global team. Despite more weekly
meetings and being in constant contact, we
seem to have moved from a unified vision
last quarter to a very disjointed and slightly
at odds vision for where to get next. I also
physical manifestation of the idea. Even

quick prototypes and experiments made our
work feel more real and tangible, and it took
us out of our own heads. Instead of getting
lost in a debate, we could not deny what we
were actually experiencing and seeing while
using the prototype or in the experiment.
I think both teams can improve how we
communicate and divide tasks, making
people at times individually responsible for
found it hard to get out of the technical

mindset of implementing our prototypes. It
was easy to be overcome and overwhelmed
by all of the technical components that
needed to get done and put off making sure
that the experience and needs for the user
are always at the forefront.
I am excited
to see where we are headed next and to

visit Finland to spend more time with our
global team.
Final Documentation
117
Spring Reflection
problem. Being able to design something
This quarter was certainly crazy, but it was
and physical show a final product at the end
nice to finally just get to focus on building out
of it all is a rewarding experience.
our ideas. As expected, we ran into many
The collaboration aspect with students from
challenges that we didnt anticipate, but I

think we handled it well. Everything seemed
to go wrong, but we kept going. I was super
impressive by everyones persistence. It
was also much easier to get things done
and make significant progress on our
project once the whole team was together.
We were able to divide and conquer. We
supported each other and made sure that if
someone needed help they got it. Overall,
I am extremely happy with how our project
turned out, despite the consistently long
nights. While not every part was designed or
made perfectly, the really challenging parts
we figured out and the entire experience we
hoped to deliver, we delivered. This class
has been a great exercise in communication,
balancing a tight budget and time schedule,
and designing something so far in the future
that the problem doesnt yet exist. I have
loved this class and I am happy to have met
and learned from my teammates.
David
another country and that have different

backgrounds has been an eye opening
experience. I have not collaborated with a
group from a different country and being able
to effectively and efficiently communicate
with each has been a struggle. A major
advantage with this collaboration is being
able to see different perspectives of the
same idea from mechanical engineers
and product design engineers have
opened up solutions that I probably would
not have noticed myself as an electrical
engineer. I have learned so much from my
team members about different areas of
engineering.
I believe that at the end of the course, I
will become a better engineer that has
experience and expertise in areas that I
probably would not have been involved in
outside of this design class. I am definitely
looking forward to the next two quarters and
being able to see all the hardwork and long
hours paid off from the final design built.
Fall Reflection
This design project has been challenging
and exciting. Since the project proposal is
set for the year 2035, being able to predict
and assume what the future will be like, in
terms of the automotive industry but also the
lifestyle of humankind, is extremely difficult.
I took this course to challenge me and be
able to apply my electrical engineering
and creative thinking to solve a wide-open
Winter Reflection
The winter quarter has been extremely
exciting. Since the fall critical function
prototype and working on the darkhorse
prototype, the direction and logic of going
from prototype to prototype has been well
defined. It has been great working on the
118
Final Documentation
prototypes although it has been very time
bonding, and set our sights to deliver a final
consuming and stressful at times. Being
design that has never been seen before,
able to see our hard work pay off and our
our team became stronger and supportive
concept starting to look more like a real
than ever. This class has taught me that
product gets me looking forward towards the
communication on all basis is one of the
spring quarter. There were many difficulties
keys to success regardless of how close or
with communication between our global
far away you are. Although I am ready to
partners but we have been more persistent
graduate from Stanford and excited to learn
in keeping each other in contact on a bi-
more from real world experiences at my
weekly basis or so. Our trip to Finland to
full time job, I will never forget what I have
converge and decide all the details on the
learned and experienced from ME310 and
final design will be challenging due to all
my teammates. My team is amazing and I
the opinions of the team members but I
am very glad that I was able to share this
am positive that we will work together and
experience with them. In the end, AudiEvolve
ultimately pull off a spectacular product to
was a success because of the hardwork,
show off at EXPE. I am more of a technical
dedicated,
person so I am definitely ready for the spring
attitudes of each of my teammates. We are
quarter to start to make our vision a reality.
AudiEvolve, transforming the journey into
supportive,
and
optimistic
the destination.
Spring Reflection
ME310 was an amazing experience. This
Goran
spring quarter was extremely intense but we
Fall Reflection
managed to complete all the goals we set. I
Before I took this course, I didnt know what
learned so much from my local teammates

as well as my global teammates. ME310
is a class that I have never experienced
before nor would I have experienced if I
did not come to Stanford. It is much better
than any typical electrical engineering class
that I have taken. I collaborated with some
of the smartest people in the world and I
was fortunate enough to learn valuable
skills from them, but also give them some
of my knowledge from past experiences.
Prior to the spring quarter, the Stanfords
team dynamics and communication with
the Aalto team was little to none. But once
we visited Finland, had some quality team
to expect. It was the first time for me to take

a course from another department and I
was really worried. This course is one year
long and if it doesnt meet my expectations,
I would be stuck there for the entire year.
Fortunately for me, so far the course was
amazing. I feel like I am learning a lot for
the first time since I came to Aalto and I
hope it will continue to be like that until the
end of the course. I have to mention that
I was positively surprised by the teaching
team, especially Harri. It seemed like he
never stops working. Whatever was the
problem, he solved it as soon as possible
and even showed personal initiative which
Final Documentation
119
is unfortunately very rare among professors.
Winter Reflection
I think it is really interesting that people who
Being a part of this course so far has been
recently finished the course are the ones
an interesting experience. Ive definitively
that are teaching you. Their knowledge is
learned a lot and I have to say that this has
really fresh and it is very easy to talk to them.
been the best course I have taken so far.
The project we got is inspiring. Thinking that
Maybe the assignments that we are doing
far ahead in the future makes you come up

with different and often shocking concepts.
It requires a lot of research in order to
understand the future and how might the
world evolve, but I always liked watching
scientific shows and science fiction movies
and this part of the project I really enjoyed.
This course also teaches you to try and
test. We did a lot of tests, but I still feel that
we should have done more. By observing
and testing we were able to identify some
problems we overlooked. It was a shame
that to this date we havent tested any of
our out of the box ideas. We had a plan to
test those ideas as soon as we come back
from our winter break.
Working in interdisciplinary team is a new
experience as well. Sometimes it was hard
to make a decision, but most of the time it
was fun. Although my teammates are not
from creative fields, in brainstorming I was
surprised how fast they relaxed and the
number of interesting ideas we got in a short
period of time surpassed all my previous
brainstorming experiences.
are not so new for me, but the their number

and intensity definitely are. Every week I am
thinking that after these few days I can take
a small brake, but it just never comes. Winter
quarter has gone very quickly. It has been
very stressful period. Almost through the
entire period we didnt know what to do. It is
still very confusing. The end is approaching
and decisions need to be made, but we
are still struggling with our focus and the
problems we are solving. I wouldnt want to
deliver something at the end that doesnt
satisfy me. The course takes eight months
and it would be a waste if I cant put the
project in the portfolio later. It means that the
concept needs to be strong and the visuals
as well. I realized through the course that
in the case we have not enough time at
the end, which might happen, that design
is going to suffer. Concept needs to work
and that is something required from us. It
needs to look appealing also at the end,
but the number of engineers and designers
in the group is so unbalanced that it will
be a struggle to convince people of the
importance of aesthetics when we approach
As a conclusion, so far this course is more
the final deadline. I am really scared that I
than a positive experience. We still have a
might not be able to show this project to
lot of work before the final presentations. I
future employers at the end.
just hope that the reaction from our client at

the end of the course will be positive as well.
Team work improved during this period.

We were equally lost when it comes to
120
Final Documentation
the concepts and ideas. That is maybe the
way to make it, but at the end it saved us a
reason why we stopped discussing and
lot money that we used to outsource cutting
arguing so much. It looks that we are all
of acrylic pieces.
scared about the final outcome of the course.
I believe this course is the most intensive,
We have now much better communication
but in the same time the best course I have
with our global team as well, but it still feels
ever taken. I strongly believe I have learned
that we are competing in a way. I know
a lot during this eight months especially in
that will improve during the spring. I am
the communication sense. I also started
looking forward to the end of the course.
doing more fast prototypes and for a
Not because I will have more free time, but
designer it is very valuable habit. Honestly
because I cant wait to see our final result.
in design part we could have done more.
I hope it will make me proud and that my
Some of the parts were really not on the
final reflection will be filled with satisfaction.
level expected for design professionals, but

in the given time frame and the scale of our
Spring Reflection
project, maybe that was the maximum we
We knew that we need to meet soon in
could have done.
order to clear all the differences between
For the end, I would like to thank everyone in
our two subteams and Stanford student visit
the teaching team for making this course so
to Finland helped us become one team and
amazing. I know they made it the way it was
to clarify our main goal and vision. It was a
and I really wish in the future the course will
very intense discussion and we had a lot
keep its culture and cult status. Also many
of disagreements but at the end we found
thanks to DF staff for helping us numerous
a compromise and we all felt excited about
times.
the final two months.
I am really proud to have been a part of
The final two months were very hard.
this great group of people. With that I mean
We worked a lot to fulfill our vision, but
everybody involved in the course and DF.
everything seemed not to go according to
I know that wherever I proceed in my life I
our plan. Simply put, whatever that could
will always come back to this year as one of
have failed, failed. At some point it looked
the best and most fun years of my life. Now
almost impossible to pull this off. At the end,
it is time for something new. Over and out!
we had to make some cuts, for example

removing windshield from our plan, but we
Sifo
managed to make all the other components.
Fall Reflection
My task in the final prototype was making
The Fall period is all about dealing with
the cabin space which included steering

wheel, dashboard and the ribbon. The
dashboard itself took more than two weeks
to built. I believe we took the worst possible
ambiguity. Our project is solving a problem

after 20 years. I feel we were walking in the
dark tunnel, expecting to see dim light from
far away. Its not as bad as it sounds like. In
Final Documentation
121
fact, once we started to observe real life, to
from learning Arduino for the robot, I found
dig into current problems, and to embrace
coding extremely interesting. The movement
new concepts, we felt we would become
of servo gave me immediate feedback on
successful future fortune tellers.
whether my code is correct or wrong. It was
As the project goes on, I have got two
my first time to code for electronics and to
biggest learning. The first is we should

prototype more other than spending time
discussing whether the idea is good or
not. I enjoy working with this team. My
favorite part is brainstorming, when crazy,
funky, weird ideas are generated from
people coming from 4 different majors and
5 different countries. Appreciating each
others ideas and building ideas upon
each others is definitely a good learning
process, and its difficult. We started from
fighting which idea we should test out, which
hindered our creativity and efficiency. The
second learning is constantly asking myself
what is the reason behind the supposed
obvious to dig out the users real need. I
didnt realize fun driving also means letting
your car reach a challenging status that you
can never achieve in real life. I didnt know
separating steering control and turning
control in different devices is actually easier
than combining them in one. Thanks to the
project, I paid more attention to daily routine
and details. The result is beneficial.
I wish in the next period, communication
between two sub teams can be more
frequent and collaboration can be more
productive. I wish we can test out more
crazy ideas and leave no regret before the
plug sensors, servos, LEDs to teensy board.

Connecting all of them in 10 days for an
amateur like me was challenging. Thanks
to the help from Design Factory staff, it
shortened my learning curve. Lesson from
this, sometimes we need to think big, with
can-do spirit.
In winter I found travel brought the team
together. When we were in Germany, the
team helped each other in trivial things.
We had to spend all time together, so we
got time to discuss something that had
never been discussed before. Running to
catch the last minute morning train, putting
on earphone whenever one was in toilet,
de-briefing in the noisy metro, amazed by
the robots and became speechless in Audi
factory, sitting anxiously to express what we
had been doing to Audi faculty, these are
valuable moments that make the team more
gathered and more open to a bold vision.
In winter I noticed communication is the key.
Especially between two global teams, we
need to spend more time on that. I noticed
our past communication strategy (hangout
once a week and one on one session once
or a week) is not enough. There are still
project ends.
things not agreed by both teams. There
Winter Refletion
through. There are important decisions
Winter period is all about learning. Started
are conflicts we two teams havent gone

we havent made together. Brainstorming
122
Final Documentation
through Google hangout was not engaging.
310 is definitely my favorite course and I
To talk about disagreement, any other
will remember the project, the people, the
communication tools than face to face cant
awesomeness forever. Im grateful for the so
fully handle the friction between two teams.
much help received from our teaching team
I look forward to having Stanford team in
and Defa people. As a business student
Helsinki. I wish well talk openly and come
Im probably the person learnt the most
up with something awesome.
in this team. Trying to find my own role in

this extremely technical project was quite a
Spring Reflection
journey. The largest learning goes to cross-
ME310 gives me the best learning
functional communication and multicultural
experience ever. Looking at our final
communication. I enjoyed realizing ideas
product, Im extremely proud of our team.
with fast prototypes, proud of getting hands
It looks just like the rendering we had one
dirty and being a workaholic. If I have to
month ago. To be honest although we aimed
say wishes, I wish we could have more user
high, I couldnt believe we can realize it in a
testing and more realistic planning of time in
generally perfect way.
design perspective. Thank you ME310! Im

ready to continue the awesomeness.
The past month was so far the busiest time

of my life. We worked around the clock
Tommi
to get things crossed from the to-do list.
Fall Reflection
The teamwork with Stanford students was
This project has been quite different from
pleasant. We supported each other whenever

whoever needed help. Although we have
two separate products, the integration
work was done very well. The cabin space
creation process was full of memories.
Although its not under critical priority, it
definitely was a critical nice to have and
affected the whole EXPE experience. We
experienced failure with the chair rotation,
frustration when dashboard cracked in the
painting stage, and desperation when the
ribbon completely broke and we almost tried
every possible way to bend it. It seems that
we had all possible failures in the last month,
but the important thing is, we overcame all of
them. Im glad that we never gave up being
persistent and creative.
the paper bike challenge. This is mainly

because of the vague brief and its million
opportunities to go for. Organizing ourselves
took a lot of time in the beginning and there
is still something that could be done better.
For example setting up internal deadlines
and plan them more precisely beforehand.
This is something that is quite different from
worklife since there is no individual who
organizes and who is responsible for all
the actions.To get working done efficently
globally is still a puzzle to solve.
So far the project has taught me mainly
social skills and teamology. Taking also
into account the cultural differences and
personal preferences has taken a bigger
picture in team working. This is quite
Final Documentation
123
different from the previous studies that
with quick and dirty prototypes basically
involve only substance learning like math
every time. Latest prototype was a good
and physics. These kinds of social skills are
example that proved that even the simplest
essential in real work life situations and it
idea takes quite a lot of time to actually put
is good to practice it in different situations
into practice and there will be most likely be
and this kind of project based course with
some unexpected obstacles along the way.
diverse team is excellent opportunity for that

This project has also made me put a lot
I personally like our vision we have at the
of thought into the future technology and
moment, even though it still needs some
future world, and I have a feeling it is really
polishing. Anticipatory sensing chair is a
going to be quite different than today. But
good concept but my opinion is that it needs
before really digging into it, I really had
a proper story around it. Freedom within the
never thought about it. What it comes to
cabin space does not add any extra value
autonomous cars, I had never even thought
for the user, if there is no reason to change
about that possibility even though I am a
position.
hard core car enthusiast (maybe that is also
Team dynamics seems to have some issues
the reason for that).
locally and globally. Somehow local working
What it comes to this documentation, I
seems to be highly inefficient and I do not
really think that content could be somewhat
know the reason why. Globally we have
different from the final documentation,
some communication issues and difficulties
especially
specification/
to agree with a common decision. We have
description part. Since we do not really
been trying to solve our local issues but
have a concept to go for, why do we have to
somehow it just does not seem to ease up.
artificially made something into that section?
What comes to the global problems, I think
Critical prototypes are only testing some of
our time together in Finland is going to be
the functions or some experiences, and
the cure for that.
the
design
therefore they are not even close to the

initial prototype.
Spring Reflection
This period was absolutely the best! Our
Winter Reflection
concept started to make sense and we were
This winter period has been even more
able to make solid common opinion about
chaotic than the fall period, at least I feel
our projects framework and what we are
so. We have been using increased amount
about to make.
of time on debating about philosophical
It was the turning point of our project, when
concepts without actually putting them into
Stanford student were allowed to come to
practice. All the ideas seem to be easy
Finland. Then we finally realized what we
to put into practice and not different and
will do, even though it took meant skipping
innovative enough. Therefore we end up
our spring break, but it was totally worth it.
124
Final Documentation
The next big event was our trip to Stanford!
misunderstandings during the course and
Then all the actual execution happened!
I have been finding myself frustrated but
Days were sick-long, but hey, we made it!
also surprised about the way our team has
The final outcome was extremely awesome
globally communicated. Skype and Google+
and it was so nice to hear all the good
with 10 hour time difference is certainly a
feedback.
challenging setting. It is easy to forget that

if you do not have the physical presence of
All in all, ME310 was the turning point of
your colleagues, the electrical representation
my life. This was easily to most demanding
is everything the counterparts sense out of
school challenge, and maybe, the most
you. That is why everybody should feed all
demanding challenge in my life! It taught
the time their doings to other members of
me a lot of stuff about human interaction,
the team. Trust becomes with understanding
attitudes and ways of thinking, that normal
and understanding starts from the knowledge
school courses cannot teach. All I wonder
you share. I have not managed to do this
now is, how will my last school year go,
by myself either. On the other hand, I tried
since it will be dramatically different than
how does it work if I feed information of our
this one.
tests to our Local ME310 Facebook group

and found a significant difference. Generally
In the end I want to thank my awesome
in our Loft there is a certain feeling that Audi
teammates, our TAs, our coaches, our
team is doing a lot of work. One piece of this
liaisons, and DF staff, without you, the magic
reputation might have been earned by active
would not have happened! You made my
communication. I wonder if we should take
year the unique ME310 experience that
this into account and start doing it within our
everybody were blabbering about in the
global team too, both ways? Once we had a
beginning. Thank you once again!
small reflective meeting together in our local

group and it felt really good to actually hear
Heikki
Fall Reflection
I have always been good at time management
and figuring out how to do things. Now, it has
all changed! I keep slipping off my deadlines
and coming to meetings unprepared. Work
is not an excuse, but it makes things harder.
I am really looking forward to the next year
where I have less things to do and more time
to focus on ME310.
The greatest learning has been anyway in
communications. There has been several
what people think in peaceful environment.

I always wonder what other people think
in order to adapt myself for the best of the
group. I really wish we can say after the
project that we have been one unite team
that succeeded in because we were bold in
communicating.
To get information of how people behave in
certain situations, observation is the most
usable tool. We tried not only to focus on
superficial behavior, but finding behavior
that repeats itself and that differs from
traditional behavior.
Final Documentation
125
Winter Reflection
discuss and be on the page of the project. I
In winter quarter we reached our first global
also wish we could build our prototype in a
communication issues. It felt that earlier

every problem was something lighter and
easy to solve by conversing. In the end of
this quarter we found two separate groups
striving for separate goals. I was acting
as a communication responsible between
global communications and there I was
moving car. Maybe there is too little time. I

hope I am wrong.
I have not heard about our corporate liaison
for several weeks now. There is definitely
room for improvement. We also will see our
Europe liaison during the spring break and
able to understand more where the other
then will be able to discuss further about our
part of the team stands. After switching the
ideas for the final port.
responsibility I was more and more lost and

it felt that less communication effort was
made. We also started using WhatsApp
program in our mobile phones, which had
more of positive effect on our group work.
We were able to see with less latency what
was happening on the other side of the
world with pictures and videos.
In our local communication there was less
organization. There was no sufficient driver
that we would have done anything especially
when Tommi was away. No todo lists were
made nor followed.
In the end I was happy to finally contribute
on technical prototype and were able to
build and code one fully functioning system.
I think it did good for our team to see how
much effort and details are to build one
single prototype until the end.
I am still not happy with our last concept
idea. It feels that is not ambitious enough
and does not cause an wow effect every
time. I am really looking forward to get the
other half of the team to Finland in order to
Hopefully we will get a new start after the

Stanfords visit and will be able to nail the
final prototype concept during the visit.
Spring Reflection
As anticipated, hard work before and
especially during the Stanford visit paid its
dividends. We came together and discussed
through our vision and felt united again. Only
levels of expectations were a bit different
among people. All in all it was clear that
there was so much to do before the whole
prototype was to be ready in the Design
EXPE in Stanford. After Stanford left we
knew what we wanted to achieve and we
started exploring different ways and ended
up trying Magneto concept and retraction
mechanism parallel. In the end we found out
that compromise between these two would
suit best to our overall concept and story.
I was responsible of making interaction
happening in the final prototype. That meant
I had to suggest and figure out what would
be suitable level of execution of working,
entertaining and future interaction for the
project, but also then make it happen. I like
126
challenges and me not having that much

of coding experience and inter-machine
communications I started took a look into
things I have liked to learn but never have
had a legitimate reason to dig into. I dont
mind hard working, I have done a quite a
bit of physical and boring stuff, which is not
challenge anymore, and now very much
enjoyed working under pressure without any
clue if I was going to succeed with the tasks
at all. All the thousands of lines of googling
and aha -moments will stay in my mind.
I was so much looking forward to come to
California and enjoy being responsible for
only one thing at the time. Before flying
over, at the same time with ME310, I was
spending nights at my work while having
part of the responsibility of our teams
performance. It was not fun at all. I have
learned that two major things I care about
seems to be maximum I can have on top of
each other.
Final Documentation
Final Documentation
127
7. Appendix
128
Final Documentation
7.1 Initial Brief
!
!
"#$!%&'()*+,$-!./&01!&2!34$/56+!789:;!1/&<$6=!2&/!=#$!>;:>?>;:9!+6+@$456!A$+/!5)!
&-!=#$!=&156!&2!+0=&-&4&0)!@/5B5-,!+-@!=#$!5-=$/+6=5&-!C$=*$$-!+0=&-&4&0)!6+/!+-@!
#04+-!@/5B$/D!!"#5)!1/&<$6=!*5''!C$!6&41'$=$@!5-!6&''+C&/+=5&-!*5=#!3+'=&!E-5B$/)5=A!5-!
F$')5-(5G!H5-'+-@D!
!
IJ=!5)!=#$!A$+/!>;9K!+-@!+0=&-&4&0)!1+))$-,$/!B$#56'$)!6+-!1/&B5@$!@/5B$/)!*5=#!
5-6/$+)$@!2/$$!=54$!+-@!+!)+2$/!@/5B5-,!$L1$/5$-6$!*#5'$!)=5''!1/&B5@5-,!=#$!&1=5&-!&2!
4+-0+'!6&-=/&'!52!=#$!@/5B$/!@$)5/$)D!!M/$+=$!+!6+C5-!=#+=!5)!!"#$%&$'%()#&*!5-!&/@$/!=&!
1/&B5@$!$-=$/=+5-4$-=!&/!*&/(5-,!)1+6$!2&/!=#$!@/5B$/!@0/5-,!+0=&-&4&0)!6&-=/&'!+-@!
)=5''!$-)0/$)%+,#%'*-.#*%+*/0+0%+,#%(&*10%&/+!$!2!/0%3/$(+-!$04!A$=!6+-!+')&!
1/&B5@$!+!2#&$0%3!*%(!$+*!))-$5%+,#%(&*!52!=#$!@/5B$/!6#&&)$)!=&!=+($!4+-0+'!
6&-=/&'DN!
!
"#$!&0=6&4$!&2!=#5)!1/&<$6=!)#&0'@!4+-52$)=!5=)$'2!+)!+!1#A)56+'!1/&=&=A1$D!
!
M&/1&/+=$!O5+5)&-!
"/$B&/!P#+--&-!
%&'()*+,$-!8'$6=/&-56)!Q$)$+/6#!O+C!
8-,5-$$/G!70'=54$@5+!311'56+=5&-)!"$+4!
6R!S;TD9T;DT:U;!
&R!UK;DTVUDS;U9!
$R!=/$B&/D)#+--&-WB*D6&4!
5/'R!K;;!M'511$/!X/5B$G!Y$'4&-=G!M3!
Final Documentation
129
7.2 Fall Brochure
Design Team
Project Background
Audi is at the forefront of innovation in automotive technology and
is dedicated to providing customers with elegant sophisticated
solutions. Autonomous vehicles (AVs) will be relatively common by
the year 2035 and Audi envisions car design will focus not only on
the driving experience, but also the riding experience. Important
areas in designing driving spaces for AVs include how people will
interact with an AV and perform activities other than driving. Even
though the future users will be comfortable with riding in an AV,
there will be times when the user will want to take control just to
enjoy the pleasure of driving.
Stanford University
Sangram Patil
Stephanie Tomasetta
David Wang
Aalto University
Goran Bjelajac
Sifo Luo
Heikki Sjman
Tommi Tuulenmki
Vision
The design team envisions a future where people will want to
regain time lost from commuting to locations where they would be
productive. The team goal is to create adaptable cabin spaces
suitable for many activites, in order to transform the journey into
the destination.
Team VW-Audi envisions driving being a secondary activity that is
performed as frequently as all the other desired activities, such as
working, relaxing, socializing, and interacting with multi-media. An
important aspect of this design would be a seamless transition
between any of these activities. The VW-Audi team will aim to
maintain the pleasure of driving while providing comfort, safety
and an easy transition that increases situational awareness.
Main Assumptions about 2035

UUsers will already trust and be comfortable with the
autonomous vehicle technology in a normal riding scenario.
Corporate Liaison
Trevor Shannon
Coach
Jeremy Dabrowiak
UDriving will not be turned over to the human driver in

emergency situations. The car will safely come to a complete
stop in those situations.
UThere will be car-to-car and car-to-infrastructure
communication.
130
Final Documentation
Left: User testing with different steering controllers. Middle: Transition Experience Prototype that utilizes a
guided matching task for gradual transfer of control. Right: Cabin Space Prototype of moving chairs.
Critical Function & Experience Prototypes

The team focused their efforts on three main areas: steering, transitioning to driving, and cabin
space. The Aalto Team created multiple prototypes to test how effective and intuitive various
steering controllers are for the driver. The Stanford Team focused on creating an experience for
gradual or direct transitioning to manual driving using a display that prompts the driver to mimic
the control actions of the AV. The intent was to increase the drivers situational awareness of the
driving scenario they are entering after being occupied by another activity. The team also tested
out a new concept for cabin space interiors that will be more suited for a wide variety of activites
by having reconfigurable chair locations.
Key Findings
UVisual representation of interactive information is too intrusive and distracting during driving
even if the visuals are projected into the driving field of view.
UPeople are better at doing multiple tasks at a time if the sources of information needed to
perform the tasks are cohesive or the same.
UWhen prompted to mimic the control actions of an AV, users didnt feel like they were driving
Design Requirements
Functional
Physical
Opportunities
U Seamless transition between any activities
UThe cabin has to be

open and have an
organized interior
U Redefine the
business model
to one that is
perhaps
subscription
based
UNon-intrusive interface to the current

activity
UKeep the user aware of the surrounding
environment and the AVs actions
UPersonalization of
the cabin space to
each user
UThe control input from the user should

maintain the pleasure of driving and be
utilized as the interaction input
UAdaptable cabin
space for doing user
desired activities
ULeverage the
social impact
of self-aware cars
Design Strategy
U Explore user interaction in the context of cabin space transitioning
UBetter define and categorize activities that would likely be performed in an AV
UIdentify emerging technologies to integrate into the cabin space based on defined activites
Final Documentation
7.3 Needfinding
To get information of how people behave in
131
Conclusions / Lessons Learned
the seat looks comfortable, users try to
certain situations, observation is the most
make themselves more comfortable. In
usable tool. We tried not only to focus on
uncomfortable chairs, users dont care
superficial behavior, but finding behavior

that repeats itself and that differs from
traditional behavior.
Cafeteria observations
As a part of our observation exercise at
Stanford University, we went to the local
cafeteria and observed people working in a
public space while eating and having drinks
at the same time. The reason behind it was
to learn about peoples working habits in a
relaxed atmosphere. (Figure 69)
Observations
One seated, people rarely change

their initial position, although their
seating position is very uncomfortable
(Figure 14: Person in uncomfortable
position)
People dont think about their comfort

in advance. They put their bags on the
table before they sit, and then once
seated, they adjust to the situation.
(They worked around the bag on the
table instead of moving it to the chair
next to them)
People on more comfortable chairs

wanted to make themselves even more
comfortable. They placed their legs on
the small table and placed laptops on
their laps.
Interior can dictate user behavior - if
about their comfort
We can manipulate users and predict

their behavior
Sometimes, users dont behave in a

logical manner
If we want people to seat comfortably,

then our design must express comfort
132
7.4 Benchmarking
Gestural steering
An Xbox Kinect Dance Revolution game
was played to learn what it would feel like to
drive using gestural commands. The setup
involved a TV, an Xbox, and the Kinect
sensor module. The sensor module had to
be placed and leveled in the correct position
to be able to capture all of the movements
of the dancer. The distance of the sensor
was also important to be able to capture
movements from head to toe. Several
dance songs were played and points were
evaluated to see how well the dancer
mimicked the animated dancers motions.
Mind Control
The prototype was set up to play two different
games. One game involved two users to
compete against each other to move the
ball towards the opponents side. The other
game involved one user to raise or lower the
ball through different ball obstacles along
the track. Buttons on the game console
manually controlled the forward movement
of the ball.
Voice Command Steering
Several different paths were taken within the
empty parking lot to exhibit a more random
driving pattern to the driver. It was essential
that the driver did not know exactly where
they were even though they were able to see
the course prior to driving. Different degrees
of voice commands were given including
hard left turn, straight, slightly right, etc
Final Documentation
Motion Sickness
The experiment set up involved a car in
which the back seat windows were covered
with black garbage bags to not allow the
passenger in the backside to see the
outside. Also garbage bags were used to
partition the front section of the car from
the back section of the car. The passenger
was reading during the tests. The driver
randomly drove to different parts of the
campus with unexpected turns and car
movements/actions.
CNC Machine Operation
The experiment set up was viewed in the
Stanford Machine Workshop. The team
observed a student that was using the
CNC machines for a project in his class/
lab section. The team first observed the
students actions with the machine without
disturbing him, and then approached him
and asked direct questions of what exactly
he is doing and why he was doing that.
Racing Game Player-to-Player Transition
The experiment involved an Xbox game
station with Forza Motorsports racing game.
One user handled controls at first, and then
the second user stepped in after the first
user relinquished controls over. Several
transitions were tested including during
turning sections, straight away sections, and
within busy clustered car sections. The case
where the second user taking over controls
without seeing what was happening prior
was also tested.
Final Documentation
133
Haptic Command Steering
Confirmation Cue Voice Indicator
Setup was the same as the voice command
The experiment used the same set up as the
steering.
motion sickness benchmark and similar to

the light indictor benchmark (without the use
Confirmation Cue Light Indictor
of the indictor console). The driver shouted
The experiment used the same set
out voice commands prior to the cars
up involved in the motion sickness
actions occurring while driving randomly to
benchmarking with an additional of an
different parts of Stanford campus.
indicator as shown in the figure below which

was made to allow the riders in the front seat
to shine a light at the indications of the car
actions. The passenger in the backseat only
had the view of the indicator. The passenger
was reading during the tests.
The first experiment tested advanced
notification of the car turning LEFT,
STRAIGHT, and RIGHT. The second
experiment tested advanced notification of
the car approaching something out of the
ordinary, SPEED BUMP, SHARP TURN, or
SUDDEN BRAKING.
(Figure 70: Light Indicator Console)
Workspace Reconfigurable Set Up

The experiment set up involved a car
dashboard, a TV that was an interactive
windshield, a movable chair on a roller
board, and a cabin space table. We had a
looping driving scene on the TV to pretend
that the user was driving, and then the user
could initiate the transition with a click of
a switch, to transition into the workspace
when desired. The user could perform any
activities within the space provided.
iOnRoad Application Android

Once a smartphone running the app
is mounted on a cars windshield or
dashboard, iOnRoad combines the visual
information collected by the smartphones
camera with GPS and accelerometer data
to provide information about the road ahead
on the smartphones display. The vehicle
in the lane ahead of the driver is displayed
and marked with a time gap in seconds,
indicating how far behind the lead vehicle
the driver is and how much time there would
be to react in an emergency. Additionally,
the road between the iOnRoad-equipped
vehicle and the car ahead is marked with
a colored overlay that goes from green for
Light Indicator Console
a safe following distance to yellow to red
134
Final Documentation
when youre following too closely. In the
Again the overlay of the image of
event that the drivers attention lapses and
the environment is very intuitive to
the vehicle ahead stops or slows suddenly,
understand
iOnRoad will flash a full-screen alert and

audio warning to grab the drivers attention
and attempt to avert an accident. iOnRoad
BMW NOTES ABOUT AUTONOMOUS
can also be set to run in the background,
DRIVING
leaving the screen free for navigation while
Track Trainer: 2007 328i BMW (estimated
it continues to keep an eye on the road (get
worth $1 mil)
it?) for potential accidents.

http://reviews.cnet.com/8301-13746_7-
Races racetracks autonomously
20067732-48.html
Technology in it
Differential GPS
(Figure 71: iOnRoad Iphone application)
Augmented for ionosphere errors

(caused by clouds, etc.) which cause it to
read length further away than they are
Good at long term data sets
Optical Lane Detection
Distance from edge of track
Camera based
Digital Map
Used for lane detection and gives
iOnRoad Iphone Application
another reference for localization

IMU & Vehicle Sensors
Observations
Once you are in a more dangerous
Gyroscopes
zone, a big yellow warning takes over
IMU is bad for racetrack because it
the display. The change in display that
is not precise enough

Good for short term data sets
dramatically is distracting.
It does not show your progress to

correct that problem.
The Car Show: Man vs. Computer Video
It is nice that it can run behind other
(Watch video here http://www.hulu.com/
apps so that you only get warnings.
watch/273766)
Must be mounted to windshield
Track Trainer driven autonomously and

by human
Conclusions
Im not driving... this is sweet.
Knowing not only the warning, but
A little disconcerting
when you have remedied the problem
Save on gas because you can have
is important
lighter cars that are electronically
Final Documentation
135
accident proof
Traffic Jam Assistant BMW I3 concept
Trains racers the lines and path of
Adaptive cruise control & steering

assistance
proper racetrack driving
Adaptive cruise control uses radar, input
A9 Autobahn: 5 series with sensors
a max speed and following distance
Technology in it:
(longitudinal control)
Controller from track trainer no
Steering Assistance lane following if
preprogrammed trajectory
there are lines- keeps car in lane (lateral
Localization
control)
Planning
(Figure 72: Drivers assistance BMW
Perception
presentation)
Highway scenario ideal because:
Dont want to drive the 4 hours in the

middle of the drive from SF to LA
High is pretty controlled
Long Strips of driving
No random variables like kids playing

in the street
Emergency Stop Assistant
Incapacitated Driver Detected (drowsy,
Drivers assistance presentation
heart attack)
Vehicle takes over
Safely moves over and stops in a

secure location
Notifies emergency services
Other Assistive Technologies:
Adaptive Brake Assistant
Night Vision with Pedestrian Detection
Traffic Jam Assistant
Remote Control Parking
Narrow Passage Assistant
Lateral Collision Avoidance
Active Hazard Braking
Camera Based Pedestrian Protection
Bigger Picture-> THE FUTURE OF HIGHLY

AUTOMATED DRIVING
Efficiency
Time- coordinated systems needed to
remove traffic (i.e. short following distance)

Fuel- predict when to brake and can
avoid those situations
Space- platooning and save drag
Convenience and Experience

Reduce Fatigue
Drive only when you want
Do something else
136
Final Documentation
Safety (feel they must justify these
Unplanned failure mode
features in terms of efficiency and
There must be enough time and cues
convenience because consumer wont
for driver to reengage
buy safety features despite wanting a
Multi Vehicle and Infrastructure
safe car)
Interaction
Reduce frequency of accidents
Mixed mode of a few autonomous
Reduce severity of accidents
cars and many human drivers

Pedestrian crossing the street
Challenges:
example
Legal Issues
Pedestrian makes eye contract with
Relevant research done by Stanford
driver to make sure they are going to
legal Fellow about autonomous
stop. What happens when there is no
driving laws
driver and the car is autonomous?
Policy spearheaded by google
Research done about how to
Nevada has strongest policy enacted
give cues back to humans about
Florida is providing liability protection
autonomous mode.
to car companies
Headlights that turn with steering has
(Figure 73)
been suggested
(Figure 74)
Legal Development of Autonomous

Vehicles
Challenges of Autonomous Vehicles
Driver Interaction
Research Sensors
Complex Interaction systems
Made by VELODYNE
How to take over at high speeds
Laser scanners
Takes time to become aware enough

to take over
Is it possible to train people to do this?
Two types of take overs
Planned
64 emitters/ receivers
Spin to get a sweeping 360 degree
detection
Data volume: 1.3 mil points/ second
How do you handle that much data?
Final Documentation
Production Sensors (Slides on following
137
page)
Elevator example
Elevators used to have a elevator man
that received calls, pulled crank, pick
you up and took you to your floor.
Now elevators are automated and you
barely notice
7.5 Technical Literature

Benchmarking
The teams research and benchmarking

efforts for existing technologies were
focused on the future of mobility, existing
autonomous and concept cars, situational
awareness, steering and control, trust and
driver psychology. A lot of research has
been carried out in the area of situational
awareness and trust related to car driving
as well as aircrafts. The team has also
explored some research articles relevant
to these areas and the key insights have
been outlined in the discussion below.
Produstion Sensor Slides
Future of mobility
(1) Personal mobility
The team looked at existing and future
Past Attempts at Autonomous Driving
technological solutions for personal
Path project1990s
mobility like the project PUMA (Personal
Autonomous platooning
Urban Mobility and Acceptability), which is
Didnt go anywhere because it relied
a collaboration between GM and Segway.
on infrastructure changes
(Reference: http://www.segway.com/
It worked by having road magnets
puma/)
embedded in the road every few feet
It values less over more; taking
to keep cars in the lanes and provide
up less space, using less energy,
feedback
produced more efficiently with fewer
IMPORTANT ASSUMPTION is that
parts, creating fewer emissions during
you can assume government support
production and operation
and you must work with infrastructure

for human drivers
The elegance and maneuverability of

dynamic stabilization combined with
138
Final Documentation
proven battery, sensing, and controls

technologies come together to solve
real transportation challenges
(2) Zipcar model versus Uber model for
car sharing
These business models lead to the

concept of collaborative consumption.
(Reference: http://www.
collaborativeconsumption.com/the-
Stanley (a mod. VW), winner of DARPAs

Grand Challenge
movement/)
Collaborative Consumption describes

the rapid explosion in traditional
sharing, bartering, lending, trading,
renting, gifting, and swapping
reinvented through network
technologies on a scale and in ways
never possible before.
It would be a new era marked by TRUST between strangers, ACCESS

instead of ownership
Boss (a mod. GM), winner of DARPAs Grand

Challenge
Existing autonomous cars and driver

assistance technologies
1. DARPA challenge
Stanley is actuated via a drive-by-wire
system developed by Volkswagen of
Americas Electronic Research Lab. The
vehicle incorporates measurements from
GPS, a 6DOF inertial measurement unit,
and wheel speed for pose estimation.
While the vehicle is in motion, the
environment is perceived through four
laser range finders, a radar system, a
stereo camera pair, and a monocular
vision system.
2. Nissan Steer by Wire technology brings

car design closer to autonomous:
http://auto.dohax.com/nissans-steerby-wire-system-brings-us-closer-toautonomouscars/
As the car encounters obstacles / another
vehicle, a pedestrian the car steers
itself to safety and stops. This technology
has already been developed. So it is
safe to assume that cars of the future will
have the built in emergency stop feature
and we only need to consider transitions
during normal driving situations or when it
is safe.
Final Documentation
139
(BMW is also working on a new

technology that detects health condition
and incapacitated drivers and stops the
car safely pulls over to the side and calls
for help).
(3) Chris Gerdes - TED Video on
autonomous cars:
http://www.youtube.com/watch?v=q1sk47
VW Hover Car
FLAmg&feature=player_embedded
Gerdes believes that the optimal
autonomous car tech will not necessarily
Trust
replace humans, but should instead act
Reference paper [3] The effects of adverse
as our coach. He believes that the optimal
condition warning system characteristics
car will combine technology with human
on driver performance: an investigation of
intuition and reflexes.
alarm signal type and threshold level

Description The study addresses the issues
Concept Cars
concerning the design of adverse condition
1. PAT Autonomous Vehicle [1]
warning systems (ACWS), which warns the
We can see that the vehicle itself has
drivers about adverse road and weather
been restructured to be like a clear
conditions, or even system conditions that
cabin with the walls themselves being
can lead to skids or accidents.
interactive interfaces.
Relevant ideas and conclusions:
2. VW hover car [2]
Results of the study indicate that drivers
The car shape is more like a pod and
respond better to a low sensitivity and
steering is a car shaped joystick. So the
graded alarm signal condition compared
controller is more intuitive and the car
to other alert configurations. Applied to
responds exactly how the user moves
this project, it means that if warnings or
around the controller.
information is too sensitive and is displayed

too often then users will not trust the system
as much and even get irritated by it. A
graded alarm system means that it goes
from a low level alarm much before to a high
level alarm when the event is near in time.
140
Final Documentation
Situational Awareness:
study of auditory warnings in aircraft
1. FORD SYNC (similar to Kia UVO (short
Relevant ideas and conclusions:
for your voice) powered by Microsoft: [4]
Pilots prefer visual over auditory
SYNC lets you use your phone,
warnings when there is enough time to
browse and choose music and find
react (Stokes et al. 1990). However if
your way to just about anywhere - all
the visual channel is overloaded there
while keeping your eyes on the road
are obvious advantages in allocating
and hands on the wheel.
some tasks to other sensory channels.
Features: Calling, Entertainment,
Advantages of auditory information:
Navigation, Assist, Vehicle Health
a) Can be received regardless of head
Report with just voice commands.
position and orientation of the user
Minimal interface and minimal buttons
b) Response time to an acoustic
to distract you.
signal
a
can
visual
be
one
shorter
than
(Kramer,
for
1994)
2. GMs OnStar 4G LTE dashboard: [6]
c) Loud enough auditory warnings
Future of automobile safety, security
attract attention regardless of the task
and infotainment
in which the user is engaged at the time
Using your smartphone to monitor
d) The user might miss visual cues in
your vehicle real time through a series
high workload environments (Edworthy,
of cameras mounted on the car
1994)
Immediate notification of any impact/
Disadvantages of auditory information:
damage transmitted directly to your
a) If the signal is too loud and
smartphone
harsh that can lead to distraction
Video tutorials combined with
b) If two alarms or signals go off at the
text based info (everything on the
same time that would be problematic
touchscreen in the car) shows how
to convey
to use the different controls (Like the
Gaver (1986) has proposed the use of
past Audi project had nfo buttons)
auditory icons. They are caricatures of
Voice command navigation.
natural sounds that describe the
Navigation system that connects with
essential feature of the event that
traffic camera allowing you to see in
they portray. For events that do not
real time the state of the traffic on the
have natural sounds, earcons have
route
been proposed which are essentially

certain sounds, which the users learn to
3. Reference paper [7] - An empirical
associate with the corresponding
Final Documentation
141
4. Reference Study [8] Collision Warning
The graphs are shown on the next page.
Design to Mitigate Driver Distraction
The black line represents the data from
This study also found that the drivers
the users and the blue line represents
trusted the graded warning system more.
the ideal data that they were expected to
The interesting conclusion was that haptic
follow. Steering and pedal data graphs
warnings were preferred to auditory
have been separated. In the graphs
warnings on several dimensions including
for the incremental transition, the three
trust, overall benefit to driving and
regions
annoyance. These results suggest that
being shown are completely autonomous
non-standard warning modes like haptic
mode, pedal matching mode and steering
cues from a vibrating seat or a vibrating
matching
control input need to be considered for
mode. In the direct transition there are
information or warning system design.
only two modes, autonomous and manual.
7.6 Critical Function

Prototype Golf Cart
7.6.1 Golf Cart CFP Results
In the incremental pedal data, once we
give them steering control along with
pedals they neglect pedals completely.
The car speed drops down a lot,
sometimes even close to a stop.
In the incremental pedal data, once

we give them steering control along
with pedals they neglect pedals
completely. The car speed drops
down a lot, sometimes even close to
a stop.
In general comparing the incremental

data versus the direct transfer data
shows that the matching error for the
entire transition sequence is much
lower in the direct transition.
It should be noted that the user is actually

given complete control of the car in the
direct transition; however the graph is
showing the time in manual mode when
the interface was visible to the user.
142
Golf Cart Transition Graph
Final Documentation
Final Documentation
7.6.2 Golf Cart C# Code for Gradual Control

using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Linq;
using System.Text;
using System.Windows.Forms;
using System.IO.Ports;
using System.IO;
// using the ADXL345 accelerometer - Sparkfun tutorial 240
// incremental
namespace WindowsFormsApplication1
{
public partial class Form1 : Form
{
public Form1()
{
InitializeComponent();
}
private void button1_Click(object sender, EventArgs e)
{
SerialPort serialPort1 = new SerialPort();
serialPort1.PortName = COM10;
serialPort1.BaudRate = 9600;
serialPort1.Open();
System.Drawing.Pen myPen;
myPen = new System.Drawing.Pen(System.Drawing.Color.Red);
myPen.Width = 2;
System.Drawing.Pen myPen_auto;
myPen_auto = new System.Drawing.Pen(System.Drawing.Color.Green);
myPen_auto.Width = 2;
143
144
Final Documentation
System.Drawing.SolidBrush myBrushGray = new System.Drawing.

SolidBrush(System.Drawing.Color.LightGray);
System.Drawing.SolidBrush myBrushYellow = new System.Drawing.
SolidBrush(System.Drawing.Color.LightBlue);
System.Drawing.Graphics formGraphics = this.CreateGraphics();
if (serialPort1.IsOpen)
textBox1.Text = Done;
else
textBox1.Text = Not;
textBox3.Font = new Font(Georgia, 16);
textBox3.BackColor = Color.LightGreen;
textBox3.BorderStyle = BorderStyle.FixedSingle;
textBox3.TextAlign = HorizontalAlignment.Center;
textBox3.Text = Throttle matched;
textBox3.Visible = false;
int trial_int = 0;
string trial_msg = 0;
int flg = 0;
while(flg == 0)
{
trial_msg = serialPort1.ReadLine();
try
{
trial_int = Int32.Parse(trial_msg);
}
catch
{
}
if (trial_int == 1234)
break;
}
Final Documentation
while (flg == 0)
{
trial_msg = serialPort1.ReadLine();
try
{
trial_int = Int32.Parse(trial_msg);
}
catch
{
}
if (trial_int == 5678)
{
flg = 1;
break;
}
}
int count = 0;
int error_count = 0;
int error = 0;
int last_error = 0;
int error_threshold = 150;
int count_threshold = 200; //should be 200
bool pedalAchieved = false;
int time_count = 0;
int label_count = 0;
progressBar1.Maximum = count_threshold;
int steering_error_count = 0;
int steering_error = 0;
int steering_last_error = 0;
int steering_error_threshold = 50;
int steering_count_threshold = 150;
bool steeringAchieved = false;
//int label_count = 0;
progressBar2.Maximum = steering_count_threshold;
145
146
Final Documentation
// dummy defines just to avoid definition errors

FileStream fs = new FileStream(Z:/Acads/Year2/ME310A/AUDI/CFP/Software/
Csharp Trials/Scenario2/defaultfile.txt,FileMode.Open,FileAccess.Read);
StreamReader sr = new StreamReader(fs);
// dummy defines just to avoid definition errors
FileStream fsw = new FileStream(Z:/Acads/Year2/ME310A/AUDI/CFP/Software/
Csharp Trials/Scenario2 + defaultfile2.txt , FileMode.Append, FileAccess.Write);
StreamWriter sw = new StreamWriter(fsw);
int pot = 0;
int pot2 = 0;
int pot_auto = 0;
int pot2_auto = 0;
int button_flg = 0;
int last_button_flg = -1;
double angle_offset = Math.PI / 2;
double angle_buffer_range = Math.PI / 5;
if (checkBox2.Checked)
{
fs = new FileStream(Z:/Acads/Year2/ME310A/AUDI/CFP/Software/Csharp Trials/
+ textBox2.Text, FileMode.Open, FileAccess.Read);
sr = new StreamReader(fs);
fsw = new FileStream(Z:/Acads/Year2/ME310A/AUDI/CFP/Software/Csharp
Trials/Scenario2/ + data + textBox2.Text, FileMode.Append, FileAccess.Write);
sw = new StreamWriter(fsw);
}
while (true)
{
//increment time counter if pedal not achieved
if (button_flg == 1)
{
time_count++;
}
Final Documentation
147
if (textBox3.Visible)
{
label_count++;
if (label_count == 150)
{
textBox3.Visible = false;
label_count = 0;
}
}
String msg0 = serialPort1.ReadLine();
String message = serialPort1.ReadLine();
String message2 = serialPort1.ReadLine();
button_flg = Int32.Parse(msg0);
pot = Int32.Parse(message);
pot2 = Int32.Parse(message2);
{
fsw = new FileStream(Z:/Acads/Year2/ME310A/AUDI/CFP/Software/Csharp
Trials/Scenario2 + textBox2.Text, FileMode.Append, FileAccess.Write);
sw = new StreamWriter(fsw);
sw.WriteLine(message + + message2);
sw.Flush();
sw.Close();
fsw.Close();
}
{
//string[] lines = System.IO.File.ReadAllLines(@Z:/Acads/Year2/ME310A/AUDI/
CFP/Software/Csharp Trials/Scenario2trial1.txt);
//string lines_check = lines[2];
String str1 = sr.ReadLine();
148
Final Documentation
pot_auto = Convert.ToInt32(str1);
pot2_auto = Convert.ToInt32(str2);
sw.WriteLine(message + + message2);
//sr.Close();
//fs.Close();
}
count++;;
if (count == 4)
{
progressBar1.Value = error_count;
progressBar2.Value = steering_error_count;
//textBox1.Text = error_count.ToString();
formGraphics.Clear(Color.FromName(Control));
//int formHeight = 370;

int formHeight = 250;
//Current Graphics
if (pedalAchieved)
{
pot2 = 100;
pot2_auto = 100;
}
int rectX1 = 0; int rectY1 = formHeight-pot2 / 3; int rectWidth = 400; int
rectHeight = 200;
int rectX1_auto = 0; int rectY1_auto = formHeight - pot2_auto / 3; int rectWidth_
auto = 400; int rectHeight_auto = 200;
// Correct the graphics to have stationary auto throttle and display error
rectY1 = rectY1 - rectY1_auto + 100;
rectY1_auto = 100;
error = Math.Abs(pot2 - pot2_auto);
steering_error = Math.Abs(pot - pot_auto);
// note down the switch
Final Documentation
149
if (checkBox2.Checked && last_button_flg == 0 && button_flg == 1)

{
sw.WriteLine(**********Switch************************);
sw.WriteLine(time_count);
sw.WriteLine(**********************************);
}
last_button_flg = button_flg;
//////////////////////////////////////////
// Pedal matching code
/////////////////////////////////////////
if (!pedalAchieved && error < error_threshold && last_error < error_threshold)
{
error_count++;
}
else if (!pedalAchieved && last_error < error_threshold && error >= error_
threshold)
{
error_count = 0;
}
last_error = error;
if (!pedalAchieved && error_count == count_threshold)
{
// successfully matched throttle pedal
int debug = 1;
pedalAchieved = true;
textBox3.Visible = true;
label_count = 0;
{
sw.WriteLine(**********Pedal Matched************************);
sw.WriteLine(**********************************);
}
150
Final Documentation
}
//////////////////////////////////////////
// Steering matching code
/////////////////////////////////////////
if (pedalAchieved && !steeringAchieved && steering_error < steering_error_
threshold && steering_last_error < steering_error_threshold)
{
steering_error_count++;
}
else if (pedalAchieved && !steeringAchieved && steering_last_error < error_
threshold && steering_error >= steering_error_threshold)
{
steering_error_count = 0;
}
steering_last_error = steering_error;
if (pedalAchieved && !steeringAchieved && steering_error_count == steering_
count_threshold)
{
// successfully matched throttle pedal
int debug = 1;
steeringAchieved = true;
textBox3.Text = Steering matched;
textBox3.Visible = true;
label_count = 0;
{
sw.WriteLine(********Steering matched**********************);
sw.WriteLine(******************************);
}
}
float centerX = (rectX1 + rectWidth / 2);
Final Documentation
151
float centerY = (rectY1 + rectHeight / 2);

pot = pot * 2 / 3 + 30 * 22 / (7 * 180);
double endPx = centerX + rectWidth * Math.Cos((pot) * Math.PI / (180) + angle_
offset) / 2;
double endPy = centerY - rectHeight * Math.Sin((pot) * Math.PI / (180) + angle_
offset) / 2;
float centerX_auto = (rectX1_auto + rectWidth_auto / 2);
float centerY_auto = (rectY1_auto + rectHeight_auto / 2);
pot_auto = pot_auto * 2 / 3 + 30 * 22 / (7 * 180);
double endPx_auto = centerX + rectWidth_auto * Math.Cos((pot_auto) * Math.
PI / (180) + angle_offset) / 2;
double endPy_auto = centerY - rectHeight_auto * Math.Sin((pot_auto) * Math.PI
/ (180) + angle_offset) / 2;
// do not start matching in autonomous mode
if (button_flg == 0)
error_count = 0;
//if (button_flg == 1)
//{
//Autonomous Graphics
{
if (pedalAchieved && !steeringAchieved)
{
// Steering related auto graphics
float start_angle = (float)((((-pot_auto) * Math.PI / (180) - angle_offset angle_buffer_range / 2) * 180 / Math.PI));
formGraphics.FillPie(myBrushYellow, (float)rectX1, (float)rectY1, (float)
rectWidth, (float)rectHeight, start_angle, (float)(angle_buffer_range * 180 / Math.PI));
}
else if(!pedalAchieved)
{
// Throttle related auto graphics
formGraphics.FillRectangle(myBrushGray, new Rectangle(rectX1_auto,
152
Final Documentation
rectY1_auto, rectWidth_auto, rectHeight_auto / 3));

formGraphics.DrawArc(myPen_auto, rectX1_auto, rectY1_auto,
rectWidth_auto, rectHeight_auto, 210, 120);
//formGraphics.DrawLine(myPen_auto, centerX, centerY, (float)endPx_
auto, (float)endPy_auto); //(float)endPx, (float)endPy
}
// }
//Current Graphics
if (!pedalAchieved)
{
// Throttle related current graphics
formGraphics.DrawArc(myPen, rectX1, rectY1, rectWidth, rectHeight, 210,
120);
}
else if (pedalAchieved && !steeringAchieved)
{
// Steering related current graphics
formGraphics.DrawLine(myPen, centerX, centerY, (float)endPx, (float)
endPy); //(float)endPx, (float)endPy
}
// reference for the passenger driver
int shift = 100;
int centerXnew = 400; int centerYnew = 200;
double endPx_auto_new = centerXnew + rectWidth_auto * Math.Cos((pot_
auto) * Math.PI / (180) + angle_offset) / 2;
double endPy_auto_new = centerYnew - rectHeight_auto * Math.Sin((pot_
auto) * Math.PI / (180) + angle_offset) / 2;
double endPx_new = centerXnew + rectWidth * Math.Cos((pot) * Math.PI /
(180) + angle_offset) / 2;
double endPy_new = centerYnew - rectHeight * Math.Sin((pot) * Math.PI /
(180) + angle_offset) / 2;
formGraphics.DrawLine(myPen, centerXnew + shift, centerYnew + shift,
(float)endPx_new + (float)shift, (float)endPy_new + (float)shift);
formGraphics.DrawLine(myPen_auto, centerXnew + shift, centerYnew + shift,
(float)endPx_auto_new + (float)shift, (float)endPy_auto_new + (float)shift); //(float)endPx,
(float)endPy
Final Documentation
}
count = 0;
}
}
myPen.Dispose();
formGraphics.Dispose();
}
private void textBox1_TextChanged(object sender, EventArgs e)
{
}
private void progressBar1_Click(object sender, EventArgs e)
{
}
private void button2_Click(object sender, EventArgs e)
{
this.Close();
}
private void numericUpDown1_ValueChanged(object sender, EventArgs e)
{
}
}
}
153
https://www.youtube.com/watch?v=reIyTHHCd-c
http://www.youtube.com/watch?v=uqeo_ZvGxE&feature=related
http://www.dailymotion.com/video/xcuin0_next-world-futuredanger-5th-april_shortfilms?search_algo=2
https://www.youtube.com/watch?
v=219YybX66MY&feature=related
https://www.youtube.com/watch?
v=K54LN9q1jSs&feature=related
http://www.youtube.com/watch?
v=7ic0JAH8rco&feature=relmfu
http://auto.howstuffworks.com/car-driving-safety/accidentshazardous-conditions/traffic.htm
Lecture about future of technology in 2030

Lecture by Michio Kaku
Speech from University professor about

the future
Another speech from the same guy
Intelligence revolution
Traffic statistics
Beyond 2000 show
Discovery channel video - part1 from the series
Future vehicles today
Just as an example of an episode of beyond 2000.It was shown 20 years

ago and we can see how much the world developed for that time.
From Tommi's link. Don't read it. It just shows how much time we spend
int raffic. It will be usefull in documentation one day.
A documentary about virtual reality, robots, AI... Interesting - It raised a
question that we can bond to a machine if it shows emotions and
personality. Maybe our car can be emotional?! Maybe then we will trust it
more?!
Discovery channel video
Basically most of the stuff introduced before gathered in one interesting

video.
This website is not 100% reliable, but it provides some vision of the
future and it has a poll on whether you will trust an autonomous
vehicle :).
Article claiming this would be in futre cars (20xx?).
Interesting techonlogy where computer adds digital reality to existing

reality. Introduced in 2007. Could maybe used on future windshield?
(20xx?).
Interesting vide of BMW using AR technology to assist repairing (2007).
Americans spend an average of 100 hours sitting in traffic every year
World in 2057
Future predictions
Future Intelligence
AR (Augmented Reality)
http://www.youtube.com/watch?v=P9KPJlA5yds
http://auto.howstuffworks.com/under-the-hood/trendsinnovations/5-future-car-technologies3.htm
http://www.youtube.com/watch?
v=U7UroLFYlzE&feature=related
Wasted time
http://www.futuretimeline.net
We might want to look into future materials like this
Human error, the most common cause

About 60-80% in cases a crash is caused by a human error.
accidents with cars.
Statistics what kinds of harm is caused by
E.g. wasted time and fuel worth about $78 billion.
traffic
Transparent materials for seats eg
Goran
Goran
Goran
Goran
Goran
Goran
Goran
Goran
Goran
Tommi
Tommi
Tommi
Tommi
Tommi
Tommi
Tommi
Sangram
Who added
Sangram
Comments
CMU also has/had a good autonomous

car program a couple of years back
Description
7.7 Online Research
Autonomous vehicle controlled with iPad
http://en.wikipedia.org/wiki/Augmented_reality
http://www.autoevolution.com/news/who-needs-parkingsensors-with-an-invisible-back-seat-50139.html
http://asasi.org/papers/2004/Shappell%20et%
20al_HFACS_ISASI04.pdf
http://auto.howstuffworks.com/car-driving-safety/accidentshazardous-conditions/traffic.htm
http://www.nytimes.
com/2011/05/29/business/economy/29view.html
No Link
Link/Title
http://ieeexplore.ieee.org/stamp/stamp.jsp?
tp=&arnumber=5945275
154
Final Documentation
Final Documentation
155
Link/Title
Description
http://en.wikipedia.org/wiki/Motion_sickness
Motion Sickness
http://www.medicinenet.com/motion_sickness/article.htm
Motion Sickness
http://www.viban.com/motionsickness.htm
http://en.wikipedia.org/wiki/Tilting_train
http://www.google.com/patents/US20040035347
Patent for motion

sickness
http://www.google.com/patents/US7717841
Link/Title
http://www.audi.co.uk/new-cars/a8/a8/safety/presense.html
http://www.audi.
com/com/brand/en/tools/advice/glossary.html
http://www.audi.
http://www.audi.
http://www.audi.
http://www.audi.
http://www.audi.
http://www.audi.
http://www.audi.
http://www.autoevolution.com/news/volvo-citysafety-lowers-insurance-rating-for-all-uk-boundv40s-50033.html
http://en.wikipedia.
org/wiki/Electronic_stability_control
http://www.autoevolution.com/news/cadillacvirtual-bumper-helps-avoid-parking-crashes49613.html
http://wot.motortrend.com/continental-wantsemergency-steer-assist-to-drive-cars-away-fromaccidents-8013.html
Motion sickness
product
Tilting Train
2006 Patent Autonomous, self
leveling, self
correcting antimotion sickness
chair, bed
Description
Audi pre sense
A solution for motion sickness
Who added
Tommi
Tommi
Steph
Active compensation
Sangram
Active compensation gyroscope based
Sangram
Look at the references. There

are many more different
solutions for motion sickness
that people have come up with.
Maybe we can just list the
interesting ones here
Sangram
Comments
Currently available system.
Includes floowing systems:
Audi adaptive cruise control,
Audi lane assist, Audi Side
Assist, Night vision, and
Audi Adaptive Lights.
Who added
Tommi
Audi adaptive cruise control
Currently available system
Tommi
Audi Adaptive Air Suspension-sport
Tommi
Audi Adaptive Lights
Tommi
Auto-dimming Mirrors
Tommi
Audi Side Assist
Tommi
Audi Parking System
Tommi
Audi lane assist
Tommi
Night vision
Tommi
AEB (Autonomous Emergency Breaking)

introduced by Volvo.
Tommi
Introduced by BMW and

Merced-Benz in 1987.
Tommi
ESC (Electric Stability Control)

ESP (Electric Stability Program)
DSC (Dynamic Stability Control)
Parking sensors mixed with autonomous parking if Currently available system

driver is about to ht something while parking.
introduced by Cadillac.
ESA (Emergency Steering Assist)
http://green.autoblog.com/2008/08/20/continentalAFFP (Accelerator Force Feedback Pedal)

introduces-force-feedback-accelerator-pedal/
http://auto.howstuffworks.com/under-thehood/trends-innovations/5-future-cartechnologies4.htm
Comments
Some brife explanation what is
motion sickness
What causes motion sickness
etc
Air Bag
In addition to breaking, the

car also knows how to steer
in emergency situation.
Introduced by continental
and it's available in near
future (2015?).
Already available.
Introduced by Continental in
2008.
Technology studied by
Mercedes-Benz. Idea is to
stop the car itself with air
bags :D.
Tommi
Tommi
Tommi
Tommi
156
Final Documentation
Link/Title
Description
Comments
99 page document - extensive discussion about the
https://cyberlaw.stanford.
White paper on legality of autonomous legality of autonomous cars (This however is not that
edu/files/publication/files/2012-Smithcars in the US
relevant now that robocars have been legalized in
AutomatedVehiclesAreProbablyLegalinTheUS_0.pdf
California)
Important points from the article and user comments:
transportation as a service model, dealership
package of owning a personal car + having free miles
on the robo-service-car, reference to transit oriented
enviromentalists/activists (we need to look at their
views and why they are against this), service
robocars solve the parking space problems, robocars
http://www.theatlanticcities.
can really blur the line between "public" and "private"
Article about autonomous cars
com/commute/2012/09/driverless-carstransportation, concept of dynamic carpooling using
reshaping the future
would-reshape-automobiles-and-transitrobocars and smartphones, problems in robotaxis in
system/3432/
handling large crowds of people after the Big Game
lets say coming out of the stadium?, City planning model of "free-floating semi-autonomous people
transporters" would only push us toward building
more disconnected places?, car owners renting out
their robocars (other more advanced models to zipcar
and car-sharing)
http://ideas.4brad.com/
Awesome blog about robocars
Leads to more links, Still being updated regularly
http://www.autoevolution.
Autonomous parking introduced by
Parking which can be controlled with a smartpohone.
com/news/nissan-hops-aboard-self-drivingNissan
Avaivable in near future (2015).
train-with-nsc-2015-video-50028.html
Link/Title
Description
http://conferences.ifpri.
Study about
org/2020conference/PDF/summary_bongaarts.pdf
popultaion growth
http://view.fdu.
Modern human
edu/files/humanvaluessustainability.
being's values
pdf
http://www.futuretimeline.
net/21stcentury/2020.htm
2020 people profile
Comments
Population of earth is likely to
increase with 3 billion between
2001-2050. Relevance to our cabin
is might not be that good :D.
Good insight in current human
values. A lot of useless stuff but
some relevant stuff also. Article
published 2010.
Completely new generation will rule
the world in future. People more like
you and me. Less religious, more
liberal...
Who added
Sangram
Sangram
Sangram
Tommi
Who added
Tommi
Tommi
Goran
Final Documentation
157
7.8 Fall Presentations
3P/33/3P
Stanford Presentation Slides
Sangram Patil
Stephanie Tomasetta
David Wang
User trust is established.
Goran Bjelajac
Sifo Luo
Heikki Sj man
Tommi Tuulenm ki

Emergency situations are responded to by the AV.
158
Final Documentation
3P/33/3P
Assumptions
The Design Space
The Design Space

Emergency situations are responded to by the AV.
Transition
Trust
Cabin
Control
Transition
Trust
Cabin
Control
Vehicle-to-Vehicle and Vehicle-to-Infrastructure

communication technology is implemented.
Transition
Testing the transition sequence from
autonomous to manual control
Transition: Gradual
Transition: Direct
_______ Display
______________ Switch
Transition: Key Insights
Visual representation of interactive information is too

intrusive and distracting during driving.

Performance is not based on the number of tasks being
performed, but the coherence between the sources of
information.
Final Documentation
159
3P/33/3P
The Design Space
Trust
Reason for trusting autonomous systems
Experience of riding in an autonomous vehicle

Performance is not based on the number of tasks being
performed, but the coherence between the sources of
information.
Transition
Trust
Cabin
People did not feel like they were driving when prompted
to mimic the AV controls.
Trust: Key Insights
Control
Trust: Key Insights
If the user sees what the cars sees, then trust is built.
Trust: Key Insights
If the user sees what the cars sees, then trust is built.
Having anticipatory cues for out of the ordinary situations
leads to reassurance that everything is working.
The Design Space
Cabin Space
Explore experience of having a moving chair in
the cabin space
Cabin Space: Key Insights
Interactive Windshield __________________
Transition
Trust
Cabin
Control
Movable Chair __________
Cabin Activity Space _______________________
Working
Relaxing
Socializing
Interacting with multimedia
Driving
160
Final Documentation
3P/33/3P
It is desirable to have the flexibility of moving around in

the cabin space.
It is desirable to have the flexibility of moving around in

the cabin space.
The Design Space
Transition
Trust
Cabin
Control
Attaching the control input to the moving chair allows for

the input to be used for multiple activities, including
driving.
Control Mechanisms
Explore possibility of a different control
mechanism that is more fun and easier to
transition
Control Mechanisms: Key Insights
Most non-conventional steering inputs are less precise

and have a steep learning curve.
Joystick Controlled Steering
Mind Control to Levitate Ball

(Mindflex Duel Game)
Design Requirements
Design Requirements
Most non-conventional steering inputs are less precise

and have a steep learning curve.
For crash-proof manual driving in the future, the control
inputs primary function is to maintain the fun of driving.
Comfortable Transition
Final Documentation
161
3P/33/3P
Design Requirements
Design Requirements
Intuitive and Non-Intrusive Interface
Design Requirements
Design Requirements
Situational Awareness
Next Steps for Audi Evolve
Maintaining the Pleasure of Driving
Maximizing Cabin Space
Organized Cabin Space
Personalized Cabin Space
Adaptable Cabin Space

Explore user interaction &
cabin space transitioning ________________
Redefine the business model

to a subscription based one ___
Identify emerging
technologies to integrate _____
Identify emerging
Define and categorize

potential activities in AV __________





162
3P/33/3P
Final Documentation

Leverage the social impact of
self-aware cars ____________________
Redefine the business model
to a subscription based one ___
Identify emerging
Questions?

Aalto Presentation Slides
Final Documentation
163
164
Final Documentation
Final Documentation
7.9 Winter Presentation
165
166
Final Documentation
Final Documentation
167
168
Final Documentation
7.10 Winter Brochure
EVOLVE
Possibilities to use the time and space in
the autonomous car are only limited by the
size of the car and peoples imagination. We
found that the most common activities people
wished to do in the car are working, relaxing,
and socializing. Every one of those activities
requires different body position and slightly
different interior arrangement. The space in
the car is very limited and there is just not
enough space to move freely. How to create
a comfortable and entertaining cabin environment is one of our tasks.
In order to ensure the quality and sophistication Audi brand presents, we need to make
driver experience during entire ride equally
interesting. That includes both the experience during activities in the car and the transition between them as well.
The transition between autonomous and
manual modes is another problem. If we look
at only the transition from driving to autonomous, we can immediately see that it is not
easy at all. Letting go the control of your life
to a machine requires much more than just

letting the steering wheel go. Drivers need a
simplified sequence of mode changing not
only to shorten their learning curve but also
to have a feeling of safety.
Vision
We envision cars in the year 2035 as moving personal space. Cars will be autonomous
and it will have two driving modes - manual
and autonomous. Drivers will no longer waste
time because of the time spent on the road,
but because of the possibility to use the time
while traveling for something else than driving. Those activities are going to be mainly
working, relaxing, and socializing. The design of the cabin will accommodate these activities as best as the size of the car allows to.
This requires us to radically restructure the
interior of the car. There might be only one
seat in front, and this driver seat will have to
be smart in a way that it senses what driver
wants and thus eases the transition between
activities and modes. Visual and audio guidance will also aid the transition by giving
enough information to the driver.
Final Documentation
169
170
Final Documentation
7.11 Dark Horse

Prototype
7.11.1 Reconfiguro - Drag and Drop
Interface Java Code
//Drag and Drop Interface
//ME310 Darkhorse Prototype
//By: VW-Audi 1/23/13
import acm.graphics.*;
import acm.program.*;
import java.awt.*;
import java.awt.event.*;
import javax.swing.*;
public class DragDrop extends GraphicsProgram implements DragDropConstants
{
private GPoint last_mouse_loc;
private GObject objectSelected;
private GRect dropArea;
private GLabel frontText;
private int rotateChair1Counter=0;
private int rotateChair2Counter=0;
private int flipBedCounter = 0;
private int rotateDeskCounter = 0;
private GCompound chairSymbol1 = new GCompound();
private GCompound chairSymbol2 = new GCompound();
private GCompound bedSymbol = new GCompound();
private GCompound tvSymbol = new GCompound();
private GCompound tableSymbol = new GCompound();
private GCompound deskSymbol = new GCompound();
private GCompound lampSymbol = new GCompound();
private GCompound cardSymbol = new GCompound();
private GCompound drinkSymbol = new GCompound();
private GLabel chairLabel;
private GLabel chairLabel2;
public GLabel bedLabel;
private GLabel deskLabel;
private GLabel lampLabel;
private GLabel tableLabel;
Final Documentation
171
private GLabel tvLabel;

private GLabel cardLabel;
private GLabel drinkLabel;
private GCompound legend = new GCompound();
public void init()
{
JButton resetAllButton = new JButton(Reset All);
JButton resetSelectedButton = new JButton(Reset Selected);
JButton rotateButton = new JButton(Rotate Selected);
add(resetAllButton, NORTH);
add(resetSelectedButton, NORTH);
add(rotateButton, NORTH);
setupDropArea();
createObjects();
createLegend();
addActionListeners();
addMouseListeners();
}
public void run()
{
}
public void setupDropArea()
{
dropArea = new GRect(DROP_AREA_WIDTH, DROP_AREA_HEIGHT);
dropArea.setLocation(DROP_AREA_CENTER_X, DROP_AREA_CENTER_Y);
add(dropArea);
frontText = new GLabel(BACK OF VAN FRONT OF VAN);
frontText.setFont(SansSerif-16);
add(frontText, DROP_AREA_CENTER_X + (DROP_AREA_WIDTH - frontText.getWidth())/2,
DROP_AREA_CENTER_Y + DROP_AREA_HEIGHT + frontText.getHeight());
}
public void createObjects()
{
createBedSymbol(bedSymbol, BED_INIT_LOC_X, BED_INIT_LOC_Y);
createChairSymbol(chairSymbol1, CHAIR1_INIT_LOC_X, CHAIR1_INIT_LOC_Y);
createChairSymbol(chairSymbol2, CHAIR2_INIT_LOC_X, CHAIR2_INIT_LOC_Y);
createTVSymbol(tvSymbol, TV_INIT_LOC_X, TV_INIT_LOC_Y);
createTableSymbol(tableSymbol, TABLE_INIT_LOC_X, TABLE_INIT_LOC_Y);
createDeskSymbol(deskSymbol, DESK_INIT_LOC_X, DESK_INIT_LOC_Y);
172
Final Documentation
createLampSymbol(lampSymbol, LAMP_INIT_LOC_X, LAMP_INIT_LOC_Y);

createDeckCardsSymbol(cardSymbol, DECK_CARDS_INIT_LOC_X, DECK_CARDS_INIT_
LOC_Y);
createDrinkSymbol(drinkSymbol, DRINK_INIT_LOC_X, DRINK_INIT_LOC_Y);
}
public void createLegend()
{
createLegendLabel();
}
public void createLegendLabel()
{
chairLabel = new GLabel(CHAIR SELECTED);
chairLabel2 = new GLabel(CHAIR SELECTED);
bedLabel = new GLabel(BED SELECTED);
tvLabel = new GLabel(TV PROJECTOR SELECTED);
tableLabel = new GLabel(LOW TABLE SELECTED);
deskLabel = new GLabel(DESK SELECTED);
lampLabel = new GLabel(LAMP SELECTED);
cardLabel = new GLabel(DECK OF CARDS SELECTED);
drinkLabel = new GLabel(DRINK SELECTED);
add(chairLabel);
add(chairLabel2);
add(bedLabel);
add(tvLabel);
add(tableLabel);
add(deskLabel);
add(lampLabel);
add(cardLabel);
add(drinkLabel);
chairLabel.setFont(SansSerif-20);
chairLabel2.setFont(SansSerif-20);
bedLabel.setFont(SansSerif-20);
tvLabel.setFont(SansSerif-20);
tableLabel.setFont(SansSerif-20);
deskLabel.setFont(SansSerif-20);
lampLabel.setFont(SansSerif-20);
cardLabel.setFont(SansSerif-20);
drinkLabel.setFont(SansSerif-20);
}
Final Documentation
173
public void createChairSymbol(GCompound chair, int xPosition, int yPosition)

{
GRect chair1 = new GRect(CHAIR_WIDTH, CHAIR_HEIGHT);
GLine line1 = new GLine(CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_WIDTH - CHAIR_
MARGIN,
CHAIR_MARGIN); //center horiz line
GLine line2 = new GLine(CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_
HEIGHT CHAIR_MARGIN); //left vert line
GLine line3 = new GLine(CHAIR_WIDTH - CHAIR_MARGIN, CHAIR_HEIGHT - CHAIR_
MARGIN, CHAIR_WIDTH
- CHAIR_MARGIN, CHAIR_MARGIN); //right vert line
chair.add(chair1);
chair.add(line1);
chair.add(line2);
chair.add(line3);
add(chair,xPosition, yPosition);
}
public void rotateChair()
{
int nextPosition = 0;
GLine line1 = null, line2=null, line3=null, line4=null, line5=null, line6=null;
GPoint objectLoc = objectSelected.getLocation();
double x = objectLoc.getX();
double y = objectLoc.getY();
GRect chair1 = new GRect(CHAIR_WIDTH, CHAIR_HEIGHT);
//Need to check which rotated position it is in
if(objectSelected == chairSymbol1)
{
nextPosition = rotateChair1Counter;
}
else if(objectSelected == chairSymbol2)
{
nextPosition = rotateChair2Counter;
}
switch(nextPosition)
{
case 0:
174
Final Documentation
//****DOWN FACING CHAIR****/

line1 = new GLine(CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_WIDTH - CHAIR_MARGIN,
CHAIR_MARGIN); //upper horiz line
line2 = new GLine(CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_HEIGHT CHAIR_MARGIN); //left vert line
line3 = new GLine(CHAIR_WIDTH - CHAIR_MARGIN, CHAIR_HEIGHT - CHAIR_MARGIN,
CHAIR_WIDTH - CHAIR_MARGIN, CHAIR_MARGIN); //right vert line
break;
case 1:
//****LEFT FACING CHAIR****/
//x1,y1 = upper_left point, x2,y2 = upper_right point
//x1,y1 = upper_right point, x2,y2 = bottom_left point
//x1,y2 = bottom_left point, x2,y2 = bottom_right point
line3 = new GLine(CHAIR_MARGIN, CHAIR_HEIGHT - CHAIR_MARGIN, CHAIR_WIDTH CHAIR_MARGIN, CHAIR_HEIGHT - CHAIR_MARGIN); //bottom horiz line
/*********/
break;
case 2:
//****UP FACING CHAIR****/
Final Documentation
175

//x1,y1 = upper_left point, x2,y2 = bottom_left point
line2 = new GLine(CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_MARGIN, CHAIR_HEIGHT
CHAIR_MARGIN); //left vert line
/******/
break;
case 3:
//*****RIGHT FACING CHAIR*****/
/*******/
176
Final Documentation
break;
default:
break;
}
{
remove(chairSymbol1);
chairSymbol1.removeAll();
chairSymbol1.add(chair1);
chairSymbol1.add(line1);
add(chairSymbol1, x, y);
}
{
remove(chairSymbol2);
chairSymbol2.removeAll();
chairSymbol2.add(chair1);
add(chairSymbol2, x, y);
}
}
public void createBedSymbol(GCompound bed, int xPosition, int yPosition)
{
GRect bed1 = new GRect(BED_WIDTH, BED_HEIGHT);
GRect pillow = new GRect(BED_WIDTH/8, BED_HEIGHT*3/4);
pillow.move(BED_WIDTH/10, pillow.getHeight()/5);
GLine sheet = new GLine(30, 0,30,BED_HEIGHT);
pillow.setFilled(true);
bed.add(bed1);
Final Documentation
bed.add(pillow);
bed.add(sheet);
add(bed, xPosition, yPosition);
}
public void flipBed()
{
GRect bed = null, pillow = null;
GLine sheet = null;
if(objectSelected == bedSymbol)
{
nextPosition = flipBedCounter;
{
case 0:
bed = new GRect(BED_WIDTH, BED_HEIGHT);
pillow = new GRect(BED_WIDTH/8, BED_HEIGHT*3/4);
pillow.move(BED_WIDTH/10, pillow.getHeight()/5);
sheet = new GLine(30, 0,30,BED_HEIGHT);
break;
case 1:
bed = new GRect(BED_WIDTH, BED_HEIGHT);
pillow = new GRect(BED_WIDTH/8, BED_HEIGHT*3/4);
pillow.move(BED_WIDTH - BED_WIDTH/10 - pillow.getWidth(), pillow.getHeight()/5);
sheet = new GLine(BED_WIDTH - 30, 0, BED_WIDTH - 30,BED_HEIGHT);
break;
default:
break;
}
remove(bedSymbol);
bedSymbol.removeAll();
bedSymbol.add(bed);
177
178
Final Documentation
bedSymbol.add(pillow);
bedSymbol.add(sheet);
add(bedSymbol, x, y);
}
}
public void createTVSymbol(GCompound tv, int xPosition, int yPosition)
{
GRect tv1 = new GRect(TV_WIDTH, TV_HEIGHT);
GOval frame = new GOval(2, 2, TV_WIDTH-5, TV_HEIGHT-5);
GLine line1 = new GLine(TV_WIDTH, TV_HEIGHT/2,TV_WIDTH + 10, TV_HEIGHT/2);
GLine line2 = new GLine(TV_WIDTH, TV_HEIGHT/2,TV_WIDTH + 10, TV_HEIGHT/2 + 10);
GLine line3 = new GLine(TV_WIDTH, TV_HEIGHT/2, TV_WIDTH + 10, TV_HEIGHT/2 - 10);
tv.add(tv1);
tv.add(frame);
tv.add(line1);
tv.add(line2);
tv.add(line3);
add(tv, xPosition, yPosition);
}
public void createTableSymbol(GCompound table, int xPosition, int yPosition)
{
GRect table1 = new GRect(TABLE_WIDTH, TABLE_HEIGHT);
GLabel tableText = new GLabel(T);
tableText.setFont(SansSerif-16);
double x = (table1.getWidth() - tableText.getWidth()/2);
double y = (table1.getHeight() + tableText.getAscent()/2);
table.add(table1);
table.add(tableText, x/2, y/2);
add(table, xPosition, yPosition);
}
public void createDeskSymbol(GCompound desk, int xPosition, int yPosition)
{
GRect desk1 = new GRect(DESK_WIDTH, DESK_HEIGHT);
GLabel deskText = new GLabel(D);
deskText.setFont(SansSerif-16);
double x = (desk1.getWidth() - deskText.getWidth()/2);
double y = (desk1.getHeight() + deskText.getAscent()/2);
desk.add(desk1);
desk.add(deskText,x/2, y/2 + deskText.getAscent()/8);
Final Documentation
add(desk, xPosition, yPosition);

}
public void rotateDesk()
{
GRect desk =null;
GLabel deskText = null;
double textX, textY;
if(objectSelected == deskSymbol)
{
nextPosition = rotateDeskCounter;
System.out.println(nextPosition);
{
case 0:
desk = new GRect(DESK_WIDTH, DESK_HEIGHT);
deskText = new GLabel(D);
textX = (desk.getWidth() - deskText.getWidth()/2);
textY = (desk.getHeight() + deskText.getAscent()/2);
add(deskText,textX/2, textY/2 + deskText.getAscent()/8);
break;
case 1:
desk = new GRect(DESK_HEIGHT, DESK_WIDTH);
deskText = new GLabel(D);
textX = (desk.getWidth() - deskText.getWidth()/2);
textY = (desk.getHeight() + deskText.getAscent()/2);
add(deskText, textX/2 - deskText.getWidth()/4, textY/2);
break;
default:
break;
}
remove(deskSymbol);
deskSymbol.removeAll();
179
180
Final Documentation
deskSymbol.add(desk);
deskSymbol.add(deskText);
add(deskSymbol, x, y);
}
}
public void createLampSymbol(GCompound lamp, int xPosition, int yPosition)
{
GOval frame = new GOval(LAMP_WIDTH, LAMP_HEIGHT);
GOval lamp1 = new GOval(LAMP_WIDTH, LAMP_HEIGHT);
lamp1.setColor(Color.YELLOW);
lamp1.setFilled(true);
lamp.add(lamp1);
lamp.add(frame);
add(lamp, xPosition, yPosition);
Audi EVOLVE March 19, 2013
169
}
public void createDeckCardsSymbol(GCompound card, int xPosition, int yPosition)
{
GRect card1 = new GRect(DECK_CARDS_WIDTH, DECK_CARDS_HEIGHT);
GLabel cardText = new GLabel(DC);
double x = (card1.getWidth() - cardText.getWidth()/2);
double y = (card1.getHeight() + cardText.getAscent()/2);
card.add(card1);
card.add(cardText,x,y);
add(card, xPosition, yPosition);
}
public void createDrinkSymbol(GCompound drink, int xPosition, int yPosition)
{
GOval drink1 = new GOval(DRINK_WIDTH, DRINK_HEIGHT);
GOval frame = new GOval(DRINK_WIDTH, DRINK_HEIGHT);
drink1.setColor(Color.BLUE);
drink1.setFilled(true);
drink.add(drink1);
drink.add(frame);
add(drink, xPosition, yPosition);
}
public void mousePressed(MouseEvent e)
{
Final Documentation
last_mouse_loc = new GPoint(e.getPoint());

objectSelected = getElementAt(last_mouse_loc);
if(objectSelected != null)
{
if((objectSelected != dropArea) && (objectSelected != frontText))
{
objectSelected.sendToFront();
{
bedLabel.setVisible(true);
bedLabel.setLocation(TEXT_ALIGN_X - bedLabel.getWidth()/2, TEXT_ALIGN_Y bedLabel.getAscent()/2);
}
else
bedLabel.setVisible(false);
{
chairLabel.setVisible(true);
chairLabel.setLocation(TEXT_ALIGN_X - chairLabel.getWidth()/2, TEXT_ALIGN_Y chairLabel.getAscent()/2);
}
else
chairLabel.setVisible(false);
{
chairLabel2.setVisible(true);
chairLabel2.setLocation(TEXT_ALIGN_X - chairLabel2.getWidth()/2, TEXT_ALIGN_Y
- chairLabel2.getAscent()/2);
}
else
chairLabel2.setVisible(false);
if(objectSelected == tvSymbol)
{
tvLabel.setVisible(true);
tvLabel.setLocation(TEXT_ALIGN_X - tvLabel.getWidth()/2, TEXT_ALIGN_Y tvLabel.getAscent()/2);
170
}
181
182
Final Documentation
else
tvLabel.setVisible(false);
if(objectSelected == tableSymbol)
{
tableLabel.setVisible(true);
tableLabel.setLocation(TEXT_ALIGN_X - tableLabel.getWidth()/2, TEXT_ALIGN_Y tableLabel.getAscent()/2);
}
else
tableLabel.setVisible(false);
{
deskLabel.setVisible(true);
deskLabel.setLocation(TEXT_ALIGN_X - deskLabel.getWidth()/2, TEXT_ALIGN_Y deskLabel.getAscent()/2);
}
else
deskLabel.setVisible(false);
if(objectSelected == lampSymbol)
{
lampLabel.setVisible(true);
lampLabel.setLocation(TEXT_ALIGN_X - lampLabel.getWidth()/2, TEXT_ALIGN_Y lampLabel.getAscent()/2);
}
else
lampLabel.setVisible(false);
if(objectSelected == cardSymbol)
{
cardLabel.setVisible(true);
cardLabel.setLocation(TEXT_ALIGN_X - cardLabel.getWidth()/2, TEXT_ALIGN_Y cardLabel.getAscent()/2);
}
else
cardLabel.setVisible(false);
if(objectSelected == drinkSymbol)
{
drinkLabel.setVisible(true);
drinkLabel.setLocation(TEXT_ALIGN_X - drinkLabel.getWidth()/2, TEXT_ALIGN_Y drinkLabel.getAscent()/2);
Final Documentation
183
}
else
drinkLabel.setVisible(false);
}
}
else
{
}
}
public void mouseDragged(MouseEvent e)
{
{
//Need to check whether if the object is the drop area, if so dont do anything and sent to
back
{
objectSelected.move(e.getX() - last_mouse_loc.getX(), e.getY() - last_mouse_loc.getY());
{
{
bedLabel.setLocation(TEXT_ALIGN_X - bedLabel.getWidth()/2,
TEXT_ALIGN_Y - bedLabel.getAscent()/2);
}
else
184
{
chairLabel.setLocation(TEXT_ALIGN_X - chairLabel.getWidth()/2,
TEXT_ALIGN_Y - chairLabel.getAscent()/2);
}
else
{
chairLabel2.setLocation(TEXT_ALIGN_X chairLabel2.getWidth()/2, TEXT_ALIGN_Y - chairLabel2.getAscent()/2);
}
else
{
tvLabel.setLocation(TEXT_ALIGN_X - tvLabel.getWidth()/2,
TEXT_ALIGN_Y - tvLabel.getAscent()/2);
}
else
{
tableLabel.setLocation(TEXT_ALIGN_X - tableLabel.getWidth()/2,
TEXT_ALIGN_Y - tableLabel.getAscent()/2);
}
else
{
deskLabel.setLocation(TEXT_ALIGN_X - deskLabel.getWidth()/2,
TEXT_ALIGN_Y - deskLabel.getAscent()/2);
}
else
Final Documentation
Final Documentation
{
lampLabel.setLocation(TEXT_ALIGN_X - lampLabel.getWidth()/2,
TEXT_ALIGN_Y - lampLabel.getAscent()/2);
}
else
{
cardLabel.setLocation(TEXT_ALIGN_X - cardLabel.getWidth()/2,
TEXT_ALIGN_Y - cardLabel.getAscent()/2);
}
else
{
drinkLabel.setLocation(TEXT_ALIGN_X - drinkLabel.getWidth()/2,
TEXT_ALIGN_Y - drinkLabel.getAscent()/2);
}
else
}
else
{
}
}
185
186
Final Documentation
else
{
objectSelected.sendBackward();
}
}
}
public void mouseClicked(MouseEvent e)
{
objectSelected = getElementAt(last_mouse_loc);
{
{
{
bedLabel.setLocation(TEXT_ALIGN_X - bedLabel.getWidth()/2, TEXT_ALIGN_Y bedLabel.getAscent()/2);
}
else
{
chairLabel.setLocation(TEXT_ALIGN_X - chairLabel.getWidth()/2, TEXT_ALIGN_Y chairLabel.getAscent()/2);
}
else
173
{
chairLabel2.setLocation(TEXT_ALIGN_X - chairLabel2.getWidth()/2, TEXT_ALIGN_Y
- chairLabel2.getAscent()/2);
}
Final Documentation
else
{
tvLabel.setLocation(TEXT_ALIGN_X - tvLabel.getWidth()/2, TEXT_ALIGN_Y tvLabel.getAscent()/2);
}
else
{
tableLabel.setLocation(TEXT_ALIGN_X - tableLabel.getWidth()/2, TEXT_ALIGN_Y tableLabel.getAscent()/2);
}
else
{
deskLabel.setLocation(TEXT_ALIGN_X - deskLabel.getWidth()/2, TEXT_ALIGN_Y deskLabel.getAscent()/2);
}
else
{
lampLabel.setLocation(TEXT_ALIGN_X - lampLabel.getWidth()/2, TEXT_ALIGN_Y lampLabel.getAscent()/2);
}
else
{
cardLabel.setLocation(TEXT_ALIGN_X - cardLabel.getWidth()/2, TEXT_ALIGN_Y cardLabel.getAscent()/2);
187
188
Final Documentation
}
else
{
drinkLabel.setLocation(TEXT_ALIGN_X - drinkLabel.getWidth()/2, TEXT_ALIGN_Y drinkLabel.getAscent()/2);
}
else
}
}
else
{
}
}
/*-----------------------*/
/*BUTTON ACTION PERFORMED*/
/*-----------------------*/
public void actionPerformed(ActionEvent e)
{
String cmd = e.getActionCommand();
if(cmd.equals(Reset All))
{
bedSymbol.setLocation(BED_INIT_LOC_X, BED_INIT_LOC_Y);
chairSymbol1.setLocation(CHAIR1_INIT_LOC_X, CHAIR2_INIT_LOC_Y);
tvSymbol.setLocation(TV_INIT_LOC_X, TV_INIT_LOC_Y);
tableSymbol.setLocation(TABLE_INIT_LOC_X, TABLE_INIT_LOC_Y);
Final Documentation
189
deskSymbol.setLocation(DESK_INIT_LOC_X, DESK_INIT_LOC_Y);
lampSymbol.setLocation(LAMP_INIT_LOC_X, LAMP_INIT_LOC_Y);
cardSymbol.setLocation(DECK_CARDS_INIT_LOC_X, DECK_CARDS_INIT_LOC_Y);
drinkSymbol.setLocation(DRINK_INIT_LOC_X, DRINK_INIT_LOC_Y);
}
else if(cmd.equals(Reset Selected))
{
{
bedSymbol.setLocation(BED_INIT_LOC_X, BED_INIT_LOC_Y);
else if(objectSelected == tvSymbol)
tvSymbol.setLocation(TV_INIT_LOC_X, TV_INIT_LOC_Y);
else if(objectSelected == tableSymbol)
tableSymbol.setLocation(TABLE_INIT_LOC_X, TABLE_INIT_LOC_Y);
else if(objectSelected == deskSymbol)
deskSymbol.setLocation(DESK_INIT_LOC_X,DESK_INIT_LOC_Y);
else if(objectSelected == lampSymbol)
lampSymbol.setLocation(LAMP_INIT_LOC_X, LAMP_INIT_LOC_Y);
else if(objectSelected == cardSymbol)
cardSymbol.setLocation(DECK_CARDS_INIT_LOC_X,
DECK_CARDS_INIT_LOC_Y);
else if(objectSelected == drinkSymbol)
drinkSymbol.setLocation(DRINK_INIT_LOC_X, DRINK_INIT_LOC_Y);
}
}
else if(cmd.equals(Rotate Selected))
{
{
{
if(++rotateChair1Counter == 4)
rotateChair1Counter = 0;
rotateChair();
190
}
{
if(++rotateChair2Counter == 4)
rotateChair2Counter = 0;
rotateChair();
}
else if(objectSelected == bedSymbol)
175
{
if(flipBedCounter == 1)
flipBedCounter = 0;
else
flipBedCounter = 1;
flipBed();
}
else if(objectSelected == deskSymbol)
{
if(rotateDeskCounter == 1)
rotateDeskCounter = 0;
else
rotateDeskCounter = 1;
rotateDesk();
}
}
}
}
Final Documentation
Final Documentation
7.12 Funky Prototype

7.12.1 Arduino Code for FSR
#include <math.h>
#define STATESWITCH 22 //Digital pin 22 (physical pin 78)
#define ONSWITCH 23 //Digital pin 23 (physical pin 77)
// Force Sensor Pins
const int B[] = {A0,A1,A2,A3};
const int T[] = {A4,A5,A6,A7};
// arrays to store the values
int BValue[] = {0,0,0,0};
int TValue[] = {0,0,0,0};
int switchState = 0;
int onSwitch = 0;
char topForceString[30];
char bottomForceString[30];
char forceString[50];
void setup()
{
//Create a serial connection to display the data on the terminal.
Serial.begin(9600);
// Set the Analog pins to inputs
for(int i=0;i<4;i++)
{
pinMode(B[i], INPUT);
pinMode(T[i], INPUT);
}
pinMode(ONSWITCH, INPUT);
pinMode(STATESWITCH, INPUT);
}
void loop(){
onSwitch = digitalRead(ONSWITCH);
//Only send/collect data when switch is on
if(onSwitch)
{
{
BValue[i] = analogRead(B[i]);
191
192
Final Documentation
TValue[i] = analogRead(T[i]);
}
//Send switch state data to matlab
Serial.println(switch state);
switchState = digitalRead(STATESWITCH);
Serial.println(switchState);
Serial.println(force data);
sprintf(topForceString,%u %u %u %u, TValue[0],TValue[1],TValue[2],TValue[3]);
sprintf(bottomForceString,%u %u %u %u, BValue[0],BValue[1],BValue[2],BValue[3]);
sprintf(forceString,%s %s, topForceString, bottomForceString);
Serial.println(forceString);
}
}
7.12.2 Matlab Data Collection Code (main.m)
clear;
close all;
name = MovingTest3;
serialPort = initialize_port();
forceCount = 1;
forceFlag = 0;
NCompare = 5;
while(1)
%Received value should be in format of attention value,meditation value
received = readSerial();
if(strncmpi(received,switch state,NCompare))
received = received(1:length(received)-2);
state = readSerial();
state = sscanf(state,%u); %%u gets rid of the \r\n
writeToFile(state,received,name);
elseif(strncmpi(received,force data,NCompare))
received = received(1:length(received)-2);
forceFlag = 1;
%Received data as a string in format (T1,T2,T3,T4,B1,B2,B3,B4)
values = readSerial();
writeToFile(values,received,name);
end
[date time] = strtok(datestr(clock));
writeToFile(time,time stamp,name);
end
Final Documentation
193
7.12.3 Matlab Data Collection Code (initialize_port.m)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%Initialize port
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function [serialPort] = initialize_port()
global serialPort
%port = /dev/tty.usbmodem1411;
port = COM13;
baudrate = 9600;
databits = 8;
stopbits = 1;
parity = none;
flowcontrol = none;
timeout = 10;
terminator = CR/LF;
%detect if port is already in use (if program does not terminate correctly)
oldSerial = instrfind(Port, port);
if(~isempty(oldSerial))
if(~strcmp(get(oldSerial(1),Status),open))
delete(oldSerial(1));
else
fclose(oldSerial(1));
delete(oldSerial(1));
end
end
%Establish new connection
serialPort = serial(port,BaudRate, baudrate, DataBits, databits, StopBits, stopbits, Parity,
parity,FlowControl, flowcontrol, Timeout,
timeout, Terminator, terminator);
fopen(serialPort);
%get(serialPort); %Check port settings
%fread(serialPort,serialPort.BytesAvailable);
return;
end
194
7.12.4 Matlab Data Collection Code (readSerial.m)

function [data] = readSerial()
global serialPort
while(1)
if(serialPort.BytesAvailable() ~= 0)
data = fscanf(serialPort);
disp(data);
break;
end
end
end
7.12.5 Matlab Data Collection Code (writeToFile.m)
function writeToFile(data, dataCategory,name)
if(strcmp(dataCategory, switch state))
fwriteID = fopen([name,/switch_state.txt],a+);
fprintf(fwriteID, %u\r\n, data);
elseif(strcmp(dataCategory, force data))
fwriteID = fopen([name,/force_data.txt],a+);
%T1,T2,T3,T4,B1,B2,B3,B4
fprintf(fwriteID, %s, data);
elseif(strcmp(dataCategory, time stamp))
fwriteID = fopen([name,/time_stamp.txt],a+);
fprintf(fwriteID, %s\r\n, data);
end
fclose(fwriteID);
end
7.12.6 Matlab Post Process Code (post_proc.m)
close all;
name = MovingTest3;
stateSwitch = textread([name,/switch_state.txt]);
forceData = textread([name,/force_data.txt]);
changeNums = [];
count = 1;
lastState = stateSwitch(1);
for i=2:1:max(size(stateSwitch))
Final Documentation
Final Documentation
if(stateSwitch(i)~=lastState)
changeNums(count) = i;
count = count+1;
end
lastState = stateSwitch(i);
end
figure
subplot(4,1,1);
plot(forceData(:,1));
plotLines(changeNums);
ylabel(Sensor T1);
title(Top Sensors);
subplot(4,1,2);
ylabel(Sensor T2);
subplot(4,1,3);
ylabel(Sensor T3);
subplot(4,1,4);
ylabel(Sensor T4);
figure
subplot(4,1,1);
ylabel(Sensor B1);
title(Bottom Sensors);
subplot(4,1,2);
ylabel(Sensor B2);
subplot(4,1,3);
ylabel(Sensor B3);
subplot(4,1,4);
195
196
Final Documentation
ylabel(Sensor B4);
saveas(figure(1),[name,/,name,top.fig])
saveas(figure(2),[name,/,name,bottom.fig])
saveas(figure(1),[name,/,name,top.jpg])
saveas(figure(2),[name,/,name,bottom.jpg])
function plotLines(changeNums)
for i=1:1:max(size(changeNums))
line([changeNums(i),changeNums(i)],[0,800],LineWidth,2,Color,Red)
end
end
7.13 Functional Prototype

7.13.1 Arduino Code - Anticipatory Chair
#include <avr.io.h>
#include <avr.interrupt.h>
#include <EEPROM.h>
#include <stdlib.h>
/*
T: Turn - Turn the chair and when back on pressure sensor turn back
D: Drive - Driving position turn off smart code and have steering wheel come out
R: Retract - Go back and recline slightly, steering wheel goes in and smart mode is on
S: Swipe
Arduino:
Recline by pushing on force sensors
Override switch
*/
//#define SMART_CODE
#define DUMB_CODE
//#define RESET_POS
#define CALIBRATE_CHAIR
// Motor control lines
#define M1_switch 4
#define M1_direc 5
#define M2_switch 6
Final Documentation
#define M2_direc 7
#define M3_switch 8
#define M3_direc 9
#define M4_switch 10
#define M4_direc 11
#define trial_pin A1
// Motor encoder input lines
#define M1_PIN 53 // PCI0 on PCINT0
#define M2_PIN 3 // INT5 interrupt
#define M3_PIN A8 // PCI2 on PCINT16
#define M4_PIN 2 // INT4 interrupt
// FSR control lines
#define MUX_BIT3 18
#define MUX_BIT2 19
#define MUX_BIT1 20
#define MUX_BIT0 21
// FSR input line
#define FSR_input A0
// M2 constants
const int M2_default = 650;
const int M2_D = 550;
const int M2_R = 950;
// M4 constants
const int M4_default = 650;
const int M4_D = 550;
const int M4_R = 800;
// Int Array to store FSR values
int FSR_values[12] = {0,0,0,0,0,0,0,0,0,0,0,0};
int M4_moveComplete = 1;
int M4_moveDirec = 0;
int M4_moveDelta = 0;
unsigned long int M4_moveStartTime = 0;
unsigned long int M4_moveStopTime = 0;
int M4_moveLastPos = 0;
int M4_moveStopDeltaTime = 800;
int M4_low_thresh = 150;
int M4_high_thresh = 675;
int M2_moveComplete = 1;
197
198
int M2_moveDirec = 0;
int M2_moveDelta = 0;
unsigned long int M2_moveStartTime = 0;
int M2_moveLastPos = 0;
int M4pos = 0; int M4lastpos = 0;
int M2pos = 0; int M2lastpos = 0;
int dumb_flg = 0;
unsigned long int trial_count = 0;
unsigned long int trial_count2 = 0;
unsigned long int codeStartTime = 0;
char last_incomingByte = l;
char incomingByte = ;
int drive = -1;
int twist = 0;
int code_disabled = 0;
int smart_code_disabled = 1;
char incoming_array[5] = { , , , , };
int incoming_count = 0;
int switchvar = 0;
unsigned long int lastTime = 0;
int forceDflg = 0;
int drive_isLast = 0;
void setup()
{
pinMode(M1_switch, OUTPUT);
pinMode(M1_direc, OUTPUT);
pinMode(M1_PIN, INPUT);
// Set MUX pins DDR
pinMode(MUX_BIT3,OUTPUT);
Final Documentation
Final Documentation
pinMode(FSR_input, INPUT);
pinMode(trial_pin, OUTPUT);
noInterrupts(); //disable interrupts
//Set Pin Change Interrupt Enable 0,1,2
PCICR |= (1<<PCIE2) | (1<<PCIE1) | (1<<PCIE0);
//Enable appropriate pin for pin change interrupt
PCMSK2 |= (1<<PCINT16);
PCMSK1 |= (1<<PCINT8);
PCMSK0 |= (1<<PCINT0); // Digital Pin 53
//Set External Interrupt 0 enable
EIMSK |= (1<<INT4) | (1<<INT5); // Pins 2 and 3
//Sense Control for EXTint0
193
EICRB |= (1<<ISC40) | (1<<ISC50); // Set it to trigger on pin change
interrupts(); //enable global interrupts
Serial.begin(9600);
Serial3.begin(9600);
codeStartTime = millis();
#ifdef RESET_POS
if(M4pos != M4_default)
moveM4(abs(M4pos-M4_default),-(M4pos-M4_default)/abs(M4pos-M4_default));
#endif
Serial.println(Init);
#ifdef CALIBRATE_CHAIR
moveM2(6000,-1);
moveM4(6000,-1);
while(M4_moveComplete != 1 || M2_moveComplete != 1)
{
M4_stallCheck();
M2_stallCheck();
//Serial.print(M4_moveComplete);
//Serial.print(\t);
199
200
Final Documentation
//Serial.println(M2_moveComplete);
}
Serial.println(Maxed Out);
EEPROM.write(0,0);
EEPROM.write(1,0);
EEPROM.write(2,0);
EEPROM.write(3,0);
Serial.println(EEPROM Updated);
M4pos = 0;
M2pos = 0;
while(M4_moveComplete != 1 || M2_moveComplete != 1)
{
M4_stallCheck();
M2_stallCheck();
//Serial.print(M4_moveComplete);
//Serial.print(\t);
}
Serial.println(Calibrated);
#else
// Read the EEPROM for last position
int M4low = EEPROM.read(0);
int M4high = EEPROM.read(1);
int M2low = EEPROM.read(2);
int M2high = EEPROM.read(3);
M4pos = word(M4high,M4low);
M2pos = word(M2high,M2low);
#endif
}
void loop()
{
if(code_disabled == 1)
{
//Serial.println(Waiting);
Final Documentation
}
else if(code_disabled == 0)
{
readFSRvalues();
EEPROM_Update(M4pos,M2pos);
// implement state machine here
if (Serial.available() > 0 || Serial3.available() > 0)
{
if(Serial.available() > 0)
{
incomingByte = Serial.read();
}
else if(Serial3.available() > 0)
{
incomingByte = Serial3.read();
}
/*
incoming_array[(incoming_count++)%5] = incomingByte;
Serial.print(incoming_array[0]); Serial.print(\t);
Serial.println(incoming_array[4]);
*/
//twist = int(incomingByte);
if(incomingByte == E)
{
EEPROM_Update(M4pos,M2pos);
code_disabled = 1;
}
else if(incomingByte == P)
{
smart_code_disabled = 1;
}
else if(incomingByte == O)
{
}
201
202
Final Documentation
/*
DELETE Later
//Redefine incoming byte to be D after some time since initiation
if(millis() - codeStartTime > 8000 && forceDflg == 0)
{
incomingByte = D;
forceDflg = 1;
Serial.print(Forcing a D); Serial.print(\t); Serial.println(last_incomingByte);
}
*/
//Redefine incoming byte to be R based on FSR inputs from steering wheel
// FSRs 4 and 5 - Steering front
// FSRs 6 and 10 - Steering back
//Serial.println(last_incomingByte);
if(drive_isLast == 1 && FSR_values[6] < 50 && FSR_values[10] < 50 && FSR_values[4] >
600 && FSR_values[5] > 600)
{
drive_isLast = 0;
incomingByte = R;
}
if(M4_moveComplete == 1 && M2_moveComplete == 1 && incomingByte != last_incomingByte && twist == 0)
{
last_incomingByte = incomingByte;
switch(incomingByte)
{
case Q: //default position reset
break;
case D:
if(M4pos != M4_D)
moveM4(abs(M4pos-M4_D),-(M4pos-M4_D)/abs(M4pos-M4_D));
if(M2pos != M2_D)
moveM2(abs(M2pos-M2_D),-(M2pos-M2_D)/abs(M2pos-M2_D));
drive = 1; // for the steering motor
Final Documentation
drive_isLast = 1; // for the new retract code

break;
case R:
if(M4pos != M4_R)
moveM4(abs(M4pos-M4_R),-(M4pos-M4_R)/abs(M4pos-M4_R));
if(M2pos != M2_R)
moveM2(abs(M2pos-M2_R),-(M2pos-M2_R)/abs(M2pos-M2_R));
drive = 0;
if(switchvar%2 == 1)
{
}
break;
case T:
//twist = 1;
break;
}
}
}
// Code for controlling steering mechanism
int timeconst = 20;
if(M4_moveComplete == 1 && M2_moveComplete == 1 && drive == 1)
{
trial_count = trial_count + 1;
if(trial_count < timeconst)
{
digitalWrite(M3_direc, HIGH);
digitalWrite(M3_switch, HIGH);
}
else if(trial_count == timeconst)
{
digitalWrite(M3_switch,LOW);
drive = -1;
trial_count = 0;
}
}
if(M4_moveComplete == 1 && M2_moveComplete == 1 && drive == 0)
203
204
Final Documentation
{
trial_count = trial_count + 1;
if(trial_count < timeconst*2)
{
digitalWrite(M3_direc, LOW);
}
else if(trial_count == timeconst*2)
{
digitalWrite(M3_switch, LOW);
drive = -1;
trial_count = 0;
}
}
// Code for controlling chair twist
// Code for controlling steering mechanism
//Serial.println(twist);
int timeconst2 = 20;
if(M4_moveComplete == 1 && M2_moveComplete == 1 && twist == 1)
{
trial_count2 = trial_count2 + 1;
if(trial_count2 < timeconst2)
{
}
else if(trial_count2 == timeconst2)
{
digitalWrite(M1_switch,LOW);
twist = 2;
EEPROM.write(4,twist);
trial_count2 = 0;
}
}
if(M4_moveComplete == 1 && M2_moveComplete == 1 && twist == 2 && FSR_values[8] >
M4_high_thresh)
{
twist = 3;
Final Documentation
}
if(M4_moveComplete == 1 && M2_moveComplete == 1 && twist == 3)
{
trial_count2 = trial_count2 + 1;
if(trial_count2 < timeconst2)
{
}
else if(trial_count2 == timeconst2)
{
twist = 0;
trial_count2 = 0;
}
}
// FSR Test code delete later!
Serial.print(FSR_values[4]); Serial.print(\t);
Serial.println(FSR_values[10]);
#ifdef DUMB_CODE
// Dumb Code
if(dumb_flg == 0)
{
//moveM2(200,1);
//moveM4(500,1);
//digitalWrite(M3_direc, HIGH);
//digitalWrite(M3_switch,HIGH);
//digitalWrite(trial_pin, HIGH);
dumb_flg = 1;
}
#endif
if(smart_code_disabled == 0)
{
// Smart Code
205
206
Final Documentation
//Serial.println(FSR_values[7]);
if(FSR_values[7] > M4_high_thresh && M4_moveComplete == 1)
{
moveM4(400,1);
//Serial.println(a);
}
else if(FSR_values[7] < M4_low_thresh && M4_moveComplete == 1)
{
moveM4(400,-1);
//Serial.println(b);
}
if(M4_moveComplete == 2 && millis()-M4_moveStopTime > M4_moveStopDeltaTime)
{
M4_moveComplete = 1;
}
//printFSRvalues();
}
// Read EEPROM values
int M4posval = EEPROM.read(0);
int M2posval = EEPROM.read(1);
//Serial.print(M4posval);
//Serial.print(\t);
//Serial.println(M2posval);
}
}
ISR(PCINT0_vect)
{
int debounceTime = 500;
if(millis()-lastTime > debounceTime)
{
switchvar = switchvar + 1;
Serial.println(switchvar);
if(switchvar%2 == 1)
{
Serial3.flush();
code_disabled = 0;
}
Final Documentation
207
else if(switchvar%2 == 0)
{
}
lastTime = millis();
if(M4_moveComplete == 2)
{
}
}
}
ISR(INT4_vect)
{
M4_moveStartTime = millis();
M4_moveLastPos = M4pos;
if(M4_moveDirec == -1)
198
M4pos = M4pos - 1;
else
M4pos = M4pos + 1;
//EEPROM_Update(M4pos,M2pos);
if(smart_code_disabled == 0)
{
if((M4_moveComplete == 0 && M4_moveDirec == 1 && FSR_values[7] < M4_low_thresh) ||
(M4_moveComplete == 0 &&
M4_moveDirec == -1 && FSR_values[7] > M4_high_thresh))
{
M4_moveStopTime = millis();
}
}
if(abs(M4pos - M4lastpos) > M4_moveDelta)
{
208
Final Documentation
}
}
ISR(INT5_vect)
{
if(M2_moveDirec == -1)
M2pos = M2pos - 1;
else
M2pos = M2pos + 1;
//EEPROM_Update(M4pos,M2pos);
if(abs(M2pos - M2lastpos) > M2_moveDelta)
{
}
}
void readFSRvalues()
{
/*for(int i=0;i<12;i++)
{
PORTD = i; FSR_values[i] = analogRead(FSR_input);
}*/int time_delay = 2;
digitalWrite(MUX_BIT3, LOW); digitalWrite(MUX_BIT2, LOW); digitalWrite(MUX_BIT1,
LOW); digitalWrite(MUX_BIT0, LOW);
delay(time_delay); FSR_values[0] = analogRead(FSR_input);
LOW); digitalWrite(MUX_BIT0, HIGH);
HIGH); digitalWrite(MUX_BIT0, LOW);
HIGH); digitalWrite(MUX_BIT0, HIGH);
digitalWrite(MUX_BIT3, LOW); digitalWrite(MUX_BIT2, HIGH); digitalWrite(MUX_BIT1,
Final Documentation
209

digitalWrite(MUX_BIT3, HIGH); digitalWrite(MUX_BIT2, LOW); digitalWrite(MUX_BIT1,
199
//Serial.println(Read done);
}
void printFSRvalues()
{
{
Serial.print(FSR_values[i]);
Serial.print(\t);
}
Serial.println( );
}
void M4_stallCheck(void)
{
//Stall check
//Serial.print(M4pos); Serial.print(\t); Serial.println(M4_moveLastPos);
210
Final Documentation
if(M4_moveComplete != 1 && (millis()-M4_moveStartTime > 2000) && (millis()-M4_moveStartTime < 4000) && (abs(M4pos M4_moveLastPos) < 100))
{
Serial.println(M4 Stall);
}
}
void M2_stallCheck(void)
{
//Stall check
Serial.print(M2pos); Serial.print(\t); Serial.println(M2_moveLastPos);
if(M2_moveComplete != 1 && (millis()-M2_moveStartTime > 2000) && (millis()-M2_moveStartTime < 4000) && (abs(M2pos M2_moveLastPos) < 100))
{
Serial.println(M2 Stall);
}
}
void moveM4(int counts, int direc)
{
M4_moveDirec = direc;
M4_moveDelta = counts;
M4lastpos = M4pos;
if(direc == 1)
{
}
else if(direc == -1)
{
Final Documentation
}
}
void moveM2(int counts, int direc)
{
M2_moveDirec = direc;
M2_moveDelta = counts;
200
M2lastpos = M2pos;
if(direc == 1)
{
}
else if(direc == -1)
{
}
}
void EEPROM_Update(int M4position, int M2position)
{
int M4high = highByte(M4position);
int M4low = lowByte(M4position);
int M2high = highByte(M2position);
int M2low = lowByte(M2position);
EEPROM.write(0,M4low);
EEPROM.write(1,M4high);
EEPROM.write(2,M2low);
EEPROM.write(3,M2high);
}
void setPwmFrequency(int pin, int divisor) {
byte mode;
if(pin == 5 || pin == 6 || pin == 9 || pin == 10) {
211
212
Final Documentation
switch(divisor) {
case 1: mode = 0x01; break;
default: return;
}
if(pin == 5 || pin == 6) {
TCCR0B = TCCR0B & 0b11111000 | mode;
} else {
}
} else if(pin == 3 || pin == 11) {
switch(divisor) {
default: return;
}
}
}
9.10.3 Kinect Code
Functional_System_Prototype.pde
import SimpleOpenNI.*;
import processing.serial.*; //import the Serial library
Serial port; // declare a new string called serial . A string is a sequence of characters (data
type know as char)
//SkeletonPoser turn;
SkeletonPoser drive;
SkeletonPoser retract;
//SkeletonPoser swipe;
SimpleOpenNI kinect;
Final Documentation
213
char state = q;
void setup() {
port = new Serial(this, /dev/tty.usbserial-A6003SBp, 9600); // initializing the object by assigning a port and baud rate (must match that
of Arduino)
size(640, 480);
kinect = new SimpleOpenNI(this);
kinect.enableDepth();
kinect.enableUser(SimpleOpenNI.SKEL_PROFILE_ALL);
kinect.setMirror(true);
// initialize the pose object
// turn = new SkeletonPoser(kinect);
drive = new SkeletonPoser(kinect);
retract = new SkeletonPoser(kinect);
// swipe = new SkeletonPoser(kinect);
/*
// rules for the right arm
turn.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.BELOW, SimpleOpenNI.
SKEL_RIGHT_ELBOW);
turn.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.RIGHT_OF, SimpleOpenNI.
SKEL_RIGHT_ELBOW);
turn.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.BELOW, SimpleOpenNI.
SKEL_RIGHT_SHOULDER);
turn.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.RIGHT_OF, SimpleOpenNI.
turn.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.BEHIND, SimpleOpenNI.
SKEL_RIGHT_ELBOW);
turn.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.BEHIND, SimpleOpenNI.
*/
drive.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.ABOVE, SimpleOpenNI.
SKEL_RIGHT_ELBOW);
drive.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.ABOVE, SimpleOpenNI.
drive.addRule(SimpleOpenNI.SKEL_LEFT_HAND, PoseRule.ABOVE, SimpleOpenNI.
SKEL_LEFT_ELBOW);
drive.addRule(SimpleOpenNI.SKEL_LEFT_ELBOW, PoseRule.ABOVE, SimpleOpenNI.
SKEL_LEFT_SHOULDER);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.BEHIND, SimpleOpenNI.
214
Final Documentation
SKEL_RIGHT_HAND);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_SHOULDER, PoseRule.BEHIND, SimpleOpenNI.SKEL_RIGHT_ELBOW);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.BELOW, SimpleOpenNI.
SKEL_RIGHT_HAND);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.ABOVE, SimpleOpenNI.
SKEL_LEFT_HAND);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.ABOVE, SimpleOpenNI.
SKEL_LEFT_ELBOW);
retract.addRule(SimpleOpenNI.SKEL_LEFT_HAND, PoseRule.BELOW, SimpleOpenNI.
SKEL_TORSO);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.RIGHT_OF, SimpleOpenNI.
SKEL_LEFT_HAND);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.BEHIND, SimpleOpenNI.
SKEL_RIGHT_HAND);
retract.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.IN_FRONT_OF, SimpleOpenNI.SKEL_RIGHT_SHOULDER);
/*
swipe.addRule(SimpleOpenNI.SKEL_RIGHT_HAND, PoseRule.LEFT_OF, SimpleOpenNI.
SKEL_LEFT_HAND);
swipe.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.LEFT_OF, SimpleOpenNI.
swipe.addRule(SimpleOpenNI.SKEL_RIGHT_ELBOW, PoseRule.IN_FRONT_OF, SimpleOpenNI.SKEL_LEFT_ELBOW);
*/
strokeWeight(5);
state = q;
}
void draw() {
background(0);
kinect.update();
image(kinect.depthImage(), 0, 0);
IntVector userList = new IntVector();
kinect.getUsers(userList);
if (userList.size() > 0) {
int userId = userList.get(0);
if( kinect.isTrackingSkeleton(userId)) {
// check to see if the user
Final Documentation
// is in the pose
/* if(turn.check(userId))
{
//if they are, set the color white
stroke(255);
state = T;
println(state);
} */
if(drive.check(userId))
{
//if they are, set the color blue
stroke(0,0,255);
state = D;
}
/* else if(retract.check(userId))
{
//if they are, set the color red
stroke(255,0,0);
state = R;
}
else if(swipe.check(userId))
{
//if they are, set the color yellow
stroke(250,250,19);
state = S;
} */
else
{
// otherwise set the color to green DEFAULT STATE
stroke(0,255,0);
state = q;
}
// draw the skeleton in whatever color we chose
drawSkeleton(userId);
port.write(state);
}
}
}
void drawSkeleton(int userId) {
215
216
Final Documentation
kinect.drawLimb(userId, SimpleOpenNI.SKEL_HEAD, SimpleOpenNI.SKEL_NECK);

kinect.drawLimb(userId, SimpleOpenNI.SKEL_NECK, SimpleOpenNI.SKEL_LEFT_SHOULDER);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_LEFT_SHOULDER, SimpleOpenNI.SKEL_
LEFT_ELBOW);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_LEFT_ELBOW, SimpleOpenNI.SKEL_LEFT_
HAND);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_NECK, SimpleOpenNI.SKEL_RIGHT_
SHOULDER);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_SHOULDER, SimpleOpenNI.SKEL_
RIGHT_ELBOW);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_ELBOW, SimpleOpenNI.SKEL_
RIGHT_HAND);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_LEFT_SHOULDER, SimpleOpenNI.SKEL_
TORSO);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_SHOULDER, SimpleOpenNI.SKEL_
TORSO);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_TORSO, SimpleOpenNI.SKEL_LEFT_HIP);
//kinect.drawLimb(userId, SimpleOpenNI.SKEL_LEFT_HIP, SimpleOpenNI.SKEL_LEFT_
KNEE);
//kinect.drawLimb(userId, SimpleOpenNI.SKEL_LEFT_KNEE, SimpleOpenNI.SKEL_LEFT_
FOOT);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_TORSO, SimpleOpenNI.SKEL_RIGHT_HIP);
//kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_HIP, SimpleOpenNI.SKEL_RIGHT_
KNEE);
//kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_KNEE, SimpleOpenNI.SKEL_
RIGHT_FOOT);
kinect.drawLimb(userId, SimpleOpenNI.SKEL_RIGHT_HIP, SimpleOpenNI.SKEL_LEFT_
HIP);
}
void drawLimb(int userId, int jointType1, int jointType2)
{
PVector jointPos1 = new PVector();
PVector jointPos2 = new PVector();
float confidence;
// draw the joint position
confidence = kinect.getJointPositionSkeleton(userId, jointType1, jointPos1);
confidence = kinect.getJointPositionSkeleton(userId, jointType2, jointPos2);
Final Documentation
217
line(jointPos1.x, jointPos1.y, jointPos1.z,

jointPos2.x, jointPos2.y, jointPos2.z);
}
// user-tracking callbacks!
void onNewUser(int userId) {
println(start pose detection);
kinect.startPoseDetection(Psi, userId);
}
void onEndCalibration(int userId, boolean successful) {
if (successful) {
println( User calibrated !!!);
kinect.startTrackingSkeleton(userId);
}
else {
println( Failed to calibrate user !!!);
kinect.startPoseDetection(Psi, userId);
}
}
void onStartPose(String pose, int userId) {
println(Started pose for user);
kinect.stopPoseDetection(userId);
kinect.requestCalibrationSkeleton(userId, true);
}
SkeletonPoser.pde
/*
pose.addRule(SimpleOpenNI.LEFT_HAND, SkeletonPoser.ABOVE, SimpleOpenNI.LEFT_
ELBOW);
pose.addRule(SimpleOpenNI.LEFT_HAND, SkeletonPoser.LEFT_OF, SimpleOpenNI.LEFT_
ELBOW);
if(pose.check(userId)){
// play the song
// with debounce
}*/
class SkeletonPoser {
SimpleOpenNI context;
ArrayList rules;
SkeletonPoser(SimpleOpenNI context){
this.context = context;
rules = new ArrayList();
218
Final Documentation
}
void addRule(int fromJoint, int jointRelation, int toJoint){
PoseRule rule = new PoseRule(context, fromJoint, jointRelation, toJoint);
rules.add(rule);
}
boolean check(int userID){
boolean result = true;
for(int i = 0; i < rules.size(); i++){
PoseRule rule = (PoseRule)rules.get(i);
result = result && rule.check(userID);
}
return result;
}
}
class PoseRule {
int fromJoint;
int toJoint;
PVector fromJointVector;
PVector toJointVector;
SimpleOpenNI context;
int jointRelation; // one of:
static final int ABOVE = 1;
static final int BELOW = 2;
static final int LEFT_OF = 3;
static final int RIGHT_OF = 4;
static final int IN_FRONT_OF = 5;
static final int BEHIND = 6;
static final int SAME_HEIGHT = 7;
PoseRule(SimpleOpenNI context, int fromJoint, int jointRelation, int toJoint){
this.context = context;
this.fromJoint = fromJoint;
this.toJoint = toJoint;
this.jointRelation = jointRelation;
fromJointVector = new PVector();
toJointVector = new PVector();
}
boolean check(int userID){
// populate the joint vectors for the user were checking
Final Documentation
context.getJointPositionSkeleton(userID, fromJoint, fromJointVector);

context.getJointPositionSkeleton(userID, toJoint, toJointVector);
int theta = 32;
fromJointVector.y = (fromJointVector.y * cos(radians(theta))) + (fromJointVector.z *
sin(radians(theta)));
fromJointVector.z = (-fromJointVector.y * sin(radians(theta))) + (fromJointVector.z *
cos(radians(theta)));
toJointVector.y = (toJointVector.y * cos(radians(theta))) + (toJointVector.z *
sin(radians(theta)));
toJointVector.z = -(toJointVector.y * sin(radians(theta))) + (toJointVector.z *
cos(radians(theta)));
boolean result;
switch(jointRelation){
case ABOVE:
result = (fromJointVector.y > toJointVector.y);
break;
case BELOW:
result = (fromJointVector.y < toJointVector.y);
break;
case LEFT_OF:
result = (fromJointVector.x < toJointVector.x);
break;
case RIGHT_OF:
result = (fromJointVector.x > toJointVector.x);
break;
case IN_FRONT_OF:
result = (fromJointVector.z < toJointVector.z);
break;
case BEHIND:
result = (fromJointVector.z > toJointVector.z);
break;
case SAME_HEIGHT:
result = (fromJointVector.y == toJointVector.y);
break;
default:
result = false;
break;
}
return result;
}
219
220
Final Documentation
7.13.2 Arduino Code - Buckle it Out

#include <Servo.h>
Servo smallMotor;
Servo largeMotor;
const int buttonPin = 2;
const int ledPin = 6;
// the number of the pushbutton pin

// Pin 6: Teensy++ 2.0 has the LED on pin 6
// variables will change:

int buttonState = 0;
// variable for reading the pushbutton status
boolean wheelStateIn = true; // state of the wheel

int pos = 40;
// This is where you change the starting and driving angle of the wheel out
const int outposition = 100;
const int inposition = 25;
// Variables
int i = 0;
void setup() {
// Connecting servos to Teensy pins
smallMotor.attach(15);
largeMotor.attach(16);
Final Documentation
// initialize the LED pin as an output:

pinMode(ledPin, OUTPUT);
// initialize the pushbutton pin as an input and activating the inner pullup resistor:
pinMode(buttonPin, INPUT_PULLUP);
// Initialize the starting position of the servos
smallMotor.write(inposition);
largeMotor.write(94);
Serial.begin(9600);
}
void outSequence()
{
// Debug command
Serial.println(OutSequence);
// Large motor sequence
for(i = 0; i < 6;i ++)
{
if(i == 0)
if(i == 5)
221
222
delay(106);
}
// Syncing time
delay(10);
// Small motor sequence
for(pos = inposition; pos < outposition; pos += 1)
{
smallMotor.write(pos);
delay(15);
}
}
void inSequence()
{
// Debug command
Serial.println(InSequence);
// Small motor sequence
for(pos = outposition; pos >= inposition; pos-=1)
Final Documentation
Final Documentation
{
smallMotor.write(pos);
delay(15);
}
// Syncing time
delay(10);
// Large motor sequence
for(i = 0; i < 5;i ++)
{
if(i == 0)
if(i > 5)
delay(110);
}
}
void loop(){
// Button bouncer while polling
223
224
Final Documentation
buttonState = digitalRead(buttonPin);
delay(10);
if (buttonState == digitalRead(buttonPin))
{
// Debug command
Serial.println(buttonState);
// Random delay
delay(100);
if (buttonState == LOW) {
// LED confirmation of button for debugging
digitalWrite(ledPin, HIGH);
if(wheelStateIn == true) {
outSequence();
wheelStateIn = false;
}
else {
}
}
Final Documentation
else {
// LED confirmation of button for debugging
digitalWrite(ledPin, LOW);
if(wheelStateIn == false)
{
inSequence();
wheelStateIn = true;
}
else {
}
}
}
// Securing motor to stop if malfunctioning
}
225
226
Final Documentation
7.14 Overview
The
The approach of Team Audi Evolve to
following quarter through prototypes that
creating a design solution began by first

getting a better understanding of the problem
statement. Understanding the existing
technology surrounding autonomous driving
and driver assistant systems helped us
to see what the challenges in this design
space truly were and what currently being
done successfully or unsuccessfully was
to overcome these challenges. This, in
addition to research into the year 2035 for
social trends and user needs, will provide
valuable insights for the designs.
Further research and investigation was
conducted in the following areas through
benchmarking, prototyping, and needfinding:
Control transitions from a human to a

machine and a machine to a human
Mechanisms for steering or controlling

a car
Confirmation cues indicating that

a vehicle or machine is properly
functioning
Motion sickness prevention
Cabin space design and use
knowledge
gained
from
these
experiences will come together in the

begin to integrate all components necessary
to design a solution that provides an
excellent riding and driving experience. The
following sections describe in greater detail
the approach the team took to understand
the problem statement and begin to tackle it.
In the following chapters, the knowledge
gained from these experiences will come
together through prototypes, integrating all
components necessary to design a solution
that provides an excellent riding and driving
experience. Previously mentioned will be
described in greater detail in needfinding
and benchmarking chapters.
Final Documentation
227
7.15 Design Reasoning

For old corporations, like Volkswagen Group,
it is important to nurture their ecosystem in
various ways. According to Growth Agenda
Limited, nurturing the ecosystem can be
divided into three different types of growth;
Core Growth, New Growth, and Emerging
Growth. These three growths are basically
separated by their differences in radicalness
and time frame.
Core
Growth
includes
incremental
innovations that do not offer dramatic

changes but are enough to keep the products
fresh. Volkswagen Group maintains core
growth by launching a new car model every
few years. (Figure 13).
Audi A3 Series
228
Final Documentation
New Growth offers substantial innovations

through new products to consumers. For
example, this was what Volkswagen Groups
Audi did by introducing the new SUV Q7 in
2005, which introduced a new type of car to
the Audi brand that they did not previously
offer. Audi has continued introducing new
vehicles in that pedigree branch (Q5 2009
and Q3 2011).
Long-term innovations that are radical and
disrupt the current industry are addressing
the Emerging Growth. This is at the center of
autonomous vehicle design and research at
Audi. Cabin space design is a crucial part of
the whole experience in an autonomous car
and it is part of Audis future investment. This
technology based innovation and looking to
the future for areas of emerging growth is
what our corporate project for Audi is about.
(Figure14)
Focus of our project
Emerging
Growth
New
Growth
Core
Growth
Focus of our project - Adapted from Growth Agenda Limited
Long term bets

Strategic ventures
Future investments
Cross sector partnerships
Business model innovation
Adjacent sector disruption
New market entry
New capabilities
Nurturing the ecosystem

New products and services
New customer acquisition
Operational enhancement
Final Documentation
229
7.16 Development Strategy
Needfinding includes finding out the users
To be able to give some direction to our
surveys, perspective tools, and information
thinking process, we needed to dig deep into

the current state of car industry, gathering
all possible data that might be useful for our
final prototype. In order to achieve this goal,
we followed Stanford Universitys design
process(Figure 15). First, we analysed the
project brief and defined the problems we
were solving. With such information we
were able to direct our needfinding and
benchmarking.
needs through observation, interviews,

search. Benchmarking on the other
hand
includes
experimenting
current
solutions and technologies. (Figure 16)

With benchmarking we try to get the best
possible knowledge of the current state of
technology so that we are able to push the
limits. With benchmarking we might also get
the golden idea how could something be
done differently.
(re) Define the Problem

Iterate to keep refining the concepts
Test
Learn from prototypes
ME310
Design Process
Needfinding and
Benchmarking
Understand the users and
design space
Prototype
Brainstorm
Rapid prototype to explore ideas
Gererate as many ideas as possible
Stanford University Design Process
230
Final Documentation
Observing
Others
Self Observation
Observation
Context
Ideal Persona
Needfinding
Interviews &
Surveys
Physical
Benchmarking
Experimental
Functional
Devision between needfinding and benchmarking

Assumptions help us to build us a picture
about the possible future world. In order to
get a feasible frame for our design work,
especially when it involves designing
something so far in the future for technology,
we have to make assumptions of the world.
These educated guesses are derived from
research, needfinding and benchmarking
we have done. Sources for these iterations
can be found among appendices (Sections
8.2 & 8.3).
Information
Search
Final Documentation
231
7.17.1 Future user
our lives and blend in so well that we cannot
While it is important to predict the state of
do without them. Specifically in terms of
the future technology and urban mobility
automotive experiences, it is already being
in general, it is also a good idea to predict
seen that the current generation views
the behavior and nature of the future user.
driving as a distraction from texting rather
The envisioned design solution needs to
than the other way around. Since being
address the needs of this future user.
connected is so easy, people want to stay

connected. This does not mean that the
The current generation has grown up in
future users would not love driving.
the technology revolution. People are

already used to the fast paced development
in technology and the effect of these
adventure would still be desirable in
new technologies on our lives

is
continually
Experiencing moments of thrill and

the future, but the essential
growing.
difference would lie in doing
Smartphones and smart
things because users
handheld devices have
want to do them rather
redefined the way we
than
interact
being forced to do them.
with
digital
media. If smart and self-aware
spending
time
cars
It is an extrapolation of
of the future are well integrated in the lives
the current scenario when
of users, we can envision that people
people are forced to drive
would show the same enthusiasm
along freeways with all the traffic
towards these autonomous machines
only because they have to travel to
and new upcoming technology as
and from work everyday. Based on
they are currently showing towards
this insight, it is highly probable that
smartphones.
users will find it desirable to have an

option to ride in a cabin space that
Internet and social networks have
is customized to their needs and the
surpassed boundaries of countries
activities that they would want to do, to
in terms of connecting people across
better utilize this time lost in commuting.
distant areas of the world. Accessing a

wide range of information has never been
The internet age has also led to more
this easy and it is going to get better in the
liberal thinking. Non-conventional work
future. There is this concept of perpetual
options are being explored. It has been
connectivity which is being predicted for
envisioned that new technology will lead to a
the future. It is only a matter of time before
great shift in working spaces, work cultures
all the existing technologies are developed
and procedures. The future users are most
to such an extent that they become a part of
likely going to work in an environment where
232
Final Documentation
physical presence is no longer required on a
7.17.2 Future infrastructure
daily basis. In such a situation and with the
increasing influence of autonomous cars, it

is highly probable that the future user is not a
Number of car will radically increase

globally
Number of cars in developed countries
very good driver without the basic assistance
will decrease and in developing
systems. The perceived completely manual
countries will raise
mode of driving is very different from the
Developed countries will shift towards
existing perception of manual driving. There
electric vehicles, but in developing
will be many assistance systems in place
countries gasoline will still be the
in the future cars. This prediction can be
primary fuel
justified on the basis of experiences of pilots
China will be the biggest market for
in airlines which were fit with autopilots and
cars. It will also be the largest car
new assistance systems in the 20th century.
manufacturer in the world
There was a time when all the pilots were
Middle east will lose its position in oil
skeptical about adopting, getting used to and
based wealth, because of the lack
trusting this technology. But currently, pilots
of interest for oil and because the oil
rely so much on this new technology that
reserves are going to expire
most of them cannot do without it.
Because the number of people will

rise on the planet, there will be huge
demand for food thus countries that are
rich in farmland will be major powers in
the future
Because of global warming, ice in

Siberia and north of Canada and
Europe will melt, which will make those
countries new world leaders in farmland
and as such the new global leaders
By 2030, transition between real and

virtual world will be complete. User will
feel and see with senses everything
what a person in virtual world sees and
feels. The technology will be already
available by 2020, but it will not be safe
and legal until 2030
More and more people will work from

home and live with parents - which
will bring a major decline in marriages.
People will have no opportunities
to meet people of the opposite sex.
Final Documentation
233
7.17.3 Futute technology
transport the users to any place they want
Augmented reality This technology is
rather than being limited to the car cabin
currently improving our perception of the
space. (Figure 18)
world by using different technologies like

sound, video, graphics, real time position
3d printing The paradigm shift from
data and haptics. In the context of the car,
copying to mass customization is offered by
it might mean that the futuristic cabin space
new technologies that can produce free form
consists of a windshield that has been
shapes with very little cost. In the future, it
transformed into an interactive interface
can be envisioned that there is automated
completely integrated into the drivers view
manufacturing by robots and 3D printers of
of the road ahead. Automotive companies
any imagined form from different materials.
have already started working on HUDs
This will have a crucial effect on everything
(Head Up Displays). The most extreme case
starting from nano products to the high
would be to use virtual reality to virtually
rise buildings and future infrastructure

development. (Figure 17)
3d Printing
234
Final Documentation
Augmented Reality
Graphene is one atom thick layer of
mater to have adjustable interior that can
graphite. It is currently the best electrical and
change shape and color on command and
heat conductor in the world and it provides
communicate with us in the same time. In
almost no electrical resistance. Also, it is
practice it means that the seat can be in a
transparent. In practice, it is predicted to
perfect shape for your comfort and at the
be essential part of all electronics in near
same time move around providing you with
future. Because it is so thin and transparent,
a nice massage. User can change interior to
the thickness and transparency of the
create a bed to sleep on or moving pieces
application depends on the material it is
can provide a treadmill to get in shape
applied on.
while traveling to work. Applications of this

technology are limitless. At this point it is still
Claytronics Is programmable matter.
under development. First prototypes used
It is constructed by having miniature
relatively large CADAMS, but new research
spherical computers CADAMS (currently
is being done with 1 mm diameter pieces.
1mm diameter size is in development)
Still, magnetic forces between pieces are not
that are connected between each other by
strong enough to support large weight, but
magnetic forces and can be programmable
scientists are hoping to improve that in the
to move and change shapes according to
near future. If the development of claytronics
our needs. First application might be a
follows the curve of development of other
3d fax machine that can send information
computer technologies, it is predicted that
with which the same 3d object is recreated
there will exist a 3d fax machine within 5
somewhere else. In the car design field, it
years.
might be possible in the future to use this
Final Documentation
235
Robotics and Artificial Inteligence
Telepresence Telepresence refers to a
Robots are getting more and more humanoid
set of technologies which allow a person
every day. Already there are few robots in
to feel as if they were present, to give the
Japan that behave like people, but their
appearance of being present, or to have
artificial intelligence is still pretty limited.
an effect, via telerobotics, at a place other
The evolution of robots is in close relation
than their true location. Today, a common
with the development of AI. The machines
application is in video conferences. In the
are learning and gaining experience through
future, it might include robots mimicking
time. According to that, it is responding to
human presence telerobotics or real time
every situation in such a way that maximizes
3D holographic images being projected
its chances of success. Autonomous cars
across the globe. In future cars, this might
are going to be just one of the applications
include having a constant company while
of this robotic and artificial intelligence
being driven somewhere or communication
technology in the future. (Figure19)
between two drivers, via telepresence.
Artificial Intelligence - is the intelligence
Humanoid Robot Azimo
of machines and robots. The machine is

learning and gaining experience through
time. According to that, it is behaving in
every situation in such a way that maximises
its chances of success.
Wireless energy transfer It is a technology
which is already common in current portable
electronics. Todays gadgets are charged
with wireless technology. The gadgets are
just placed on a pad which is plugged in, and
they start charging. In the future, wireless
technology can impact other markets,
including car industry. Parking spaces can
be charging stations, cars can be without
wires, everything inside of a car can be
automatically charged and powered. There
is even the possibility of having dynamic
wireless charging of car batteries.
236
7.17.4 Ideal Future Persona

The Audi project aims for users in year 2035,
at which time Audis customer base will have
a big shift from now. From the demographic
perspective, US market is nearly saturated,
and China is likely to become the next
largest car consuming country. From the
consumer behavior perspective, the new
consumer generation is one born in comfort,
well educated, and familiar with IT since
childhood. In addition, different from Audis
current business which mainly focuses on
personal cars, the future business might be
driven by the mature autonomous driving
technology and step into car sharing area.
Instead of owning a car, customers would
be able to order a car online that allows
great levels of customization. During the
fall quarter, the team has considered future
users as the car owners, but next steps
would involve developing the persona
further based on predicted subscription
based business models.
Final Documentation
Final Documentation
237
The first persona was created under

conservative imagination. This persona
was highly affected by our current view of
the world, attitudes and stereotypes of the
typical Audi users at the moment. Tommy
Yuppie, a 35 year-old American investment
banker works in Wall Street, New York. Hes
married, and has a 5 year-old girl. In free
time, he likes to play golf, work out, and
party with friends. Living outside of the city,
he has to drive to work every morning. He
always wants to use time efficiently instead
of spending it on waiting. Thats why he
needs the self-driving car to drive him in
rush hour so that he can do conference
calls, prepare meetings, or take a nap.
(Figure 20 - top)
The second persona integrates our advanced

future predictions. Based on those, our ideal
user is a 43-year-old Chinese man named
Howard Huo. He works in a hospital doing
nano medic research. He is divorced, and
has one child living with her Mom. He is
well paid from work, loves to spend money
eating in restaurants with colleagues and
friends. He uses the car to locate the best
restaurant in Guangzhou. Sometimes when
he gets drunk, and needs the car to take
him home safely. Every Saturday, he picks
his daughter up from her Moms. They drive
outside the city to enjoy nature and by then
Howard will take over the manual control.
(Figure 20 - bottom)
First ideal persona (top)
Second ideal persona (bottom)
238
Final Documentation
To get a perspective of the challenge, which

is handed to us, the team began by gathering
7.18 Needfinding
information with different tools introduced by
To get the best possible understanding of
Bill Cockayne. First, the team used a tool
our viable future users, the team did a lot of
called Context Map. The purpose of this tool
different types of needfinding. Needfinding
is to start figuring out the central concerns in
is an exercise of understanding and
the corporate project brief. After realizing it
building empathy for the target user group
was the cabin of the car, the team identified
by conducting interviews and ethnographic
eight different stakeholders involved in the
studies. This is in the core of ME310 design
ideal Autonomous Cars Cabin as seen in
process of understanding the problem. This
the Figure below. In conclusion we came
section introduces the key discoveries..
rI
se
U
Auto
Piloted
Car Cabin
ks
pa
Cars
or
W
Drivers
s
er ce
riv n
D ista
s
As
ce
Infrastructure
nt
Si
er
n
io
ot ss
M cne
fa
ce
Users
up with a graph shown in Figure 21 below.
Final Documentation
7.18.1 Context Map

To get a perspective of the challenge, which
is handed to us, the team began by gathering
information with different tools introduced at
239
Motion sickness is
highly subjective
Stanford lectures. First, the team used a tool

called Context Map. The purpose of this tool
is to start figuring out the central concerns in
the corporate project brief. After realizing it
Observations
sick relatively quickly (about after 5
was the cabin of the car, the team identified

eight different stakeholders involved in the
ideal Autonomous Cars Cabin as seen in
minutes)
moving vehicles before but this
with a graph shown in Figure below.
experiment expedited it
processing: feeling the movement of
gaining first hand experience in several key
the bus, but not seeing the movement
areas described in the design development

be the user has provided important insights
into the design space.
This may be caused by the

incongruous information the brain is
The team began the design process by
overview. Understanding what it feels like to
The person doing the test had

experienced motion sickness in
the Figure below. In conclusion we came up
7.18.2 Self-Observation
Team member started to feel motion
Sight can not be blocked totally.
People in a moving vehicle need

to get some visual signals or
No Visuals Experiment
confirmation of the movement
The idea behind this experiment was

to observe how it feels to be in public
No Sound Experiment
transportation without being able to see
The team wanted to experiment with blocking
anything. A team member sat in the middle

seat of the bus with eyes covered and
closed. The team member remained awake
for the duration of the experience. The bus
ride was about an hour long through curvy
roads. The purpose of this experiment was
to see if the removal of the outside view from
our future cabins would affect future users
and it what ways.
sound and see if it had an effect on the user

like blocking sight did. For this experiment,
the test person put noise cancelling plugs
on and had a normal one-hour-long bus ride
while sitting in the middle seat of the bus.
The test person was able to see everything
during the whole journey.
240
Final Documentation
Observations
Noise cancellation did not have a
Visual awareness proved to be the most
dramatic effect since the test person
crucial considering the motion sickness
did not feel motion sick at all
in this case
Test
person
was
more
focused
Even a little clue of motion once in
visually to outside environment since
awhile, prevents the feeling of motion
phonological sensations did not distract
sickness
test persons vigilance

Conclusions/Lessons Learned
Normal Trip In a Bus II
This test was done the same way as Normal
Blocking the sound does not negatively

impact motion sickness
Improving concentration on visuals may

actually helps prevent motion sickness
Trip In a Bus I (different activities involved),

but now the team member sat in the back of
the bus in order to determine if seat location
impacted motion sickness.
Normal Trip In a Bus I

The experiment was done to test the
Observations
hypothesis that people get more motion sick
In this case, the test person started to
when sitting in the back of the bus by first
feel minor motion sickness when just
testing how the user feels when performing
sitting in the rear part of the bus
activities while sitting in the front of the bus.

The team member began their bus ride
sitting in the front of the bus doing different
Motion sickness is highly subjective,
activities such as reading, listening music,
some people get motion sick more
playing with smartphone, and using laptop.
easily. This hypothesis might get some

1. Reason for feeling motion sick just
Observations
Listening to music and being aware

of the environment did not cause any
sense of motion sickness
Reading, playing with smartphone, and

using laptop caused a minor feeling of
motion sickness after about 15 minutes
The feeling of motion sickness eased

almost
immediately
insurance from our survey
after
outside the bus windows
looking
sitting in the rear part of the bus

might be caused by several different
reasons for example:
1. There is not that much room around
the passenger (wall directly behind)
2. Sense of acceleration is greater in
turn (longer radius to wheels)
3. Worse air ventilation
Final Documentation
241
Normal Trip In a Bus III
7.18.3 Interviews
This setup differed a bit from first two Normal
Interviews are key in acquiring personal data
Trips. The team member sat backwards in
from users and beginning to empathize with
the front part of the bus (i.e. with his back
them. The team tried to interview people
facing the direction of motion). This was to
that have experienced similar situations
test the impact of travelling while facing the
to driving in an autonomous vehicle. Also
opposite direction of motion.
speaking to futurists, future enthusiastic
Observation
insights into the motivations and needs of
The team member did not experience

motion sickness for the majority of the
users, and experts provide even more

the Audi user in 2035.
trip, except slightly at the end
Bus interviews
Even though the passenger has
Riding in an autonomous car can resemble
awareness of motion, the passenger
the experience of being driven in public
was not able to predict and prepare for
transportation. Since there is no need for
upcoming turns
driving, users can focus on doing something
else. During daily commutes, people can
The fact that passenger cannot predict
dedicate their time to whatever they need.
what is going to happen might be one of
In order to learn about current habits during
the reasons for motion sickness
commutes, the team interviewed people

commuting on trains, buses and metros in
Helsinki. (Figure 22)
242
Final Documentation
Keynotes
driving slightly worse, but is necessary

in order to drive at low velocity
Many current cars turn off the power
steering in high velocity
Drivers like to hold the bottom part of
of the travel, they tend to better utilize
the steering wheel to get a better feeling
their time on public transport.
of the road
People organize their time better if it is

a longer trip. (a trip between cities)
If there is no need for transferring or if

they are familiar with the exact duration
On the track, power steering makes
transportation; usually they listen to

music, read or just sit and think.
People rarely do work on public
The steering wheel is the heart and

soul of racecar driving
One passenger said that she doesnt

drive often and would feel more
confused and insecure with all the driver
assistance systems
Every time she drives, she thinks about

the safety of her dog in the back.
One girl thought it was boring to watch
The steering wheel is

the heart and soul of
racecar driving
the same scenery every day while going

to work
Many people consider traveling time as
Car Laboratory interview
time to rest
The Aalto team visited the Car Laboratory

in Aalto University to speak with experts
Car enthusiast I, II, and III
involved in research and innovations in

the automotive industry. Learning what
Gaining insights into what makes car
directions they see the automotive industry
enthusiasts not only love Audis, but love
moving in 25 years and what aspects are
driving. Racecar drivers and other car
important to consider when designing future
enthusiasts were the teams target users
vehicles. The interviewee was already
for a series of interviews. These users pay
familiar with similar works in this area and
attention to certain details that a casual driver
he was very knowledgeable. He showed
disregards or might not even appreciate.
the team a presentation explaining in detail

everything he thought was important in
Key Notes
Car dynamics (e.g. aerodynamics of the

car) are really important in high speed
driving
In order to be able to drive on a track,

the hands must be able to cross over
each other
designing a car for the future. (Figure 23)
Final Documentation
243
CarLab Interview
Key Notes
comes from hands on the wheel the
as expected. Explosion of the airbag
fastest (0.1 - 0.3 sec), proceed by the
is very powerful and often injures the
inner ear (balance) and sight.
Passive safety will be overtaken by
The extreme ends of drivers will be
active safety (If autonomous driving
younger and older than today, because
is so safe, there will be no need for
autonomous driving will allow it. Together
seatbelts and airbags any more)
with new user demographics, comes a
Feedback from the car movements
Passive safety in the car wasnt as safe
passenger.
Cultural
considerations
must
be
certain change in driving habits.
understood in designing for different
Steering with feet might be a more
regions - western cultures are individual
intuitive way to steer, because it
and eastern cultures are more family
resembles already learned walking
oriented.
movements
244
Figure 25 : PIlot Interviews
Final Documentation
Final Documentation
245
Pilot interviews I and II

To understand the transition experience
from autonomous mode to manual mode
and manual mode to autonomous mode, the
team went to interview pilots in a private jet
company called Jetflite Ltd. The team also
interviewed amateur pilot who flies planes
as a hobby. The team hoped to get some
insight into how it feels to hand over the
controls to a machine and to be onboard
when the machine is in control. (Figure 25)
Key Notes
Autopilot is more like a cruise control,

pilot has to be aware of the situation
There are two pilots in aprivate jet cabin

just to make sure there is always at least
one human who is aware of what the
plane is doing
What does airplane autopilot do?

Function 1: Controls trim tab and
stabilizes the plane (basic).
Function 2: Keeps the plane on certain
height
Function 3: Flying by predefined route
after take-off
Autopilot is turned on and off by pressing

a button
Usually autopilot is not used for takeoffs

and landings
When plane is in autopilot mode, the

pilot is occupied checking the weather,
estimating landing times, checking
gasoline conditions, etc.
Pilots imagined certain situations
246
Final Documentation
where the autonomous functions of
Pilot Instructor at Palo Alto
the car are the most usable: truck
Regional Airport
driver gets tired during long-distance
The pilot instructor showed the team the
travels, in traffic jams, parking, and

for preventing human errors (semiautonomous)

In order to ensure human trusts

the autopilot, the machine creates
situational awareness by providing
regular information updates (current
speed, gas status, distance to
destination, weather, etc).
To make the transition between

autopilot and manual mode as easy
as possible, the driver should decide
exactly when they wish to take over
cockpit of a Cessna airplane (two-seater).

The experience in a small aircraft with
autopilot activated is very similar to the
concept of an autonomous vehicle driving.
The team was able to get the experience
of what kinds of activities the pilots perform
in the cabin space while autopilot is
engaged, and the different kinds of steering
mechanisms used. (Figure 26)
Key Notes

joystick and yoke steering (personal
control.

beforehand. Alert can be sound,

seat vibration, color change. E.G.
If the car is about to run out of gas,
the car should remind the driver of a
mandatory stop in the next closest
gas station.

preference - both of them are very
Passive transition: in case the

system goes wrong, pilot is alerted
It takes time for the pilot to

synchronize manual steering with
the previous plane movements from
the autopilot. Especially in windy
weather when the plane has trimmed
itself to fly against the wind, the pilot
has to be able to adjust his steering
to the weather conditions during the
transition.
Different steering mechanisms -
sensitive)

Autopilot modes - three - ascending/

descending, waypoints, altitude
Auto pilot landing required during bad

weather
Semi Auto pilot - guidance system

which you manually follow
Pilots use smart phones and ipads in

the cockpit; they dont always have
hands on the controls.
Final Documentation
247
Cockpit of Cessna Plane
No jerk occurs in transitioning to

manual control
There is limited space surrounding the

yoke.
Layered and redundant safety

mechanisms exist to ensure safety
In the flight autopilot, pilots perform

a precautionary action - maintaining
hands near the control- during the
transition from manual mode to
autopilot
Conclusions
On longer trips, pilots perform activities like
reading and playing games on electronic
devices when in autopilot. Having a
comfortable and flexible cabin space is
important to be able to enjoy the riding
experience. On the transition side, the
interface with which the pilot interacts should
be more intuitive, non-redundant, and more
user friendly.
248
Final Documentation
7.18.4 EMT
(Emergency Medical Technician)
They have medic catchers for secondary

safety. Though there are seatbelts for
The team was able to view the backend of an
primary safety, these seatbelts are not
ambulance at the Stanford Hospital. EMTs
that safe and often not used since
have to work in the back of these vehicles
they allow the medic to reach across
while being transported to the hospital in
the patient to the other side of the
a fast manner, while performing medical
ambulance.
procedures on patients. (Figure 27)
The height of the passenger cabin in the

vehicle is also very important for more
Key Notes
workspace and easy accessibility into
and out of the ambulance for medics.
He noted that it never gets comfortable

being in the back of an ambulance.
They just deal with the discomfort.
Stability is an issue especially when
Having a workspace that is flexible,
performing delicate tasks like giving an
organized, and comfortable is important
IV.
especially when performing delicate
He told us there was no real solution for
tasks. A space that is designed for
stabilizing IV administration; they just do
many activities instead of one particular
their best with what they have.
activity is key.
Accessibility of equipment is necessary
Shelves
EMT Vehicle Interior
Patient Bed
Seat Belts
Final Documentation
249
7.18.5 Survey
To discover users potential needs while
driving or riding a car, the team created a
survey about the current drivers driving
habits, their favorite activities inside a
vehicle, and unpleasant moments they have
experienced as a driver or a passenger.
The survey has 9 questions with no direct
indication of autonomous driving. It was sent
out to our Facebook friends.
(Figure 28: Servey Results)
Results

62 responses. 77% of them are in the

age group 18-26. 21% of them are in
the age group 27-35. 39% are female.
61% are male.
Females
of:
enjoy
privacy,
driving
easy
because
door-to-door
transportation, music, good scenery,

and hanging out with friends.

Males enjoy driving because of: enjoy

their own time, enjoy driving at the limits
like speed and drifting, feeling of control
Some crazy things they would like to

do in the car: massage chairs, stretch,
sleep, watch a movie, hang out with
people in the back seat, prepare food,
read, text while driving, shower, surfing
the internet, and gaming.
60% of survey takers get motion sick

while reading in the back of the car and
on a bus. More than 80% of survey
takers never get motion sick while
reading in a train, metro, or plane.
250
Final Documentation
What is your age?
What is your gender?
18 - 26 (18)
Male (38)
27 - 35 (13)
36 - 45 (1)
Female (24)
18 - 26
48
77%
27 - 35
13
21%
27 - 35
2%
Female
24
39%
Male
38
61%
How often do you feel motion sick while reading in the following vehicles - Metro?
Always
Sometimes
Always
5%
Sometimes
10%
Hardly
53
85%
Hardly
0
11
22
33
44
55
How often do you feel motion sick while reading in the following vehicles - Plane?
Always
Sometimes
Hardly
0
11
22
33
44
55
Always
0%
Sometimes
15%
Hardly
53
85%
Final Documentation
251
How often do you feel motion sick while reading in the following vehicles - Car front seat?
Always
Sometimes
Always
15%
Sometimes
13
21%
Hardly
40
65%
Hardly
0
16
24
32
40
How often do you feel motion sick while reading in the following vehicles - Car back seat?
Always
Always
15
24%
Sometimes
Sometimes
22
35%
Hardly
25
40%
Hardly
0
10
15
20
25
How often do you feel motion sick while reading in the following vehicles - Bus?
Always
Sometimes
Always
15%
Sometimes
27
44%
Hardly
26
42%
Hardly
0
10
15
20
25
30
How often do you feel motion sick while reading in the following vehicles - Train?
Always
Sometimes
Hardly
0
10
20
30
40
50
Always
3%
Sometimes
10
16%
Hardly
50
81%
252
Final Documentation
Phone calling & texting
61
98%
People want more comfortable setup
Internet surfing through...
49
79%
and more space in the car
Listening to music
61
98%
People want more communication and
Mobile gaming
38
61%
engagement with other passengers.
Reading
42
68%
People want to experience dangerous
Working (laptop or paper)
29
47%
driving mode without sacrificing safety.
Chatting with friends
55
89%
People want to use their time more
Eating and drinking beverage
55
89%
efficiently in long driving trip (multi-
Drinking alcohol
27
44%
tasking)
Changing clothes
39
63%
People dont want to drive in situations
Putting on make-up
14
23%
like: boring scenery, long waiting,
Other
18
29%
parking
As a passenger, what kind of thing have you done inside a vehicle?
Phone calling & texting

Internet surfing through mobile phone
Listening to music
Mobile gaming
Reading
Working (laptop or paper)
Chatting with friends
Eating and drinking beverage
Drinking alcohol
Changing clothes
Putting on make-up
Other
0
12
24
36
48
60
72
People may select more than one checkbox, so percentages may add up to more than 100%
Final Documentation
253
7.19 Benchmarking
transition, steering, motion sickness, and
From the brainstorming sessions that the
confirmation cues. The team identified that
team conducted, there were many areas
the areas of interests could be separated
of interest that needed to be explored for
into three categories:
benchmarking. Benchmarking is an exercise

of exploring as many areas as possible
1. Physical Steering and transition
that relate to the problem statement, such
2. Psychological Trust and confirmation
as technology research, predictions, etc.
cues
Along with needfinding these two form
3. Cabin/Environment Motion sickness
the base of design. The areas areas for
and workspace
benchmarking included: trust, workspace,
Trust
Confirmation
Cues
Wokspace
Benchmarking
Motion
Sickness
Transition
Steering
254
Final Documentation
7.19.1 Steering Benchmarking
Observations
The team investigated various methods that
between the dancer and the person
could be used as steering mechanisms for

vehicles. The methods include: gestural,
mind control, voice commands, and haptic
mimicking the moves of the dancers
and these ways of steering mechanisms

could possibly show a more intuitive and
effective way to transition from autonomous
to manual mode instead of using the
traditional steering wheel.
Gestural Steering Benchmark
In order to understand what it would feel like
to drive using gestural commands, the team
tried out an Xbox Kinect dancing game to
see if gestures would be efficient to control.
(Figure 30)
Same concept applies semi-autonomous

features which has a guide to follow in
feedback. The steering benchmarking was

defined as a physical aspect of the design
There was noticeable reaction time
order to help maneuver
Initial training and warm up is needed

to get acclimated to the system
Need to have some built in error

allowance so the user has a better
chance at mimicking the commands
Conclusions/Insights
There are several insights that came from
benchmarking a gesture-based system. A
gesture based steering system can work if
the reaction time between the gesture and
response is minimized for best performance.
There also would be a learning curve
associated with this system. A training mode
can be implemented as a guide to follow
system to help the users get familiar with the
system and gain more confidence, but need
Gestural Steering Benchmark
to compensate if the user cannot mimic the

actions precisely.
Mind Control Benchmark

One non-obvious steering mechanism
that the team brainstormed was the use
of mind control. Being able to steer using
brainwaves and mind control is advanced,
but has the possibility to eliminate any
physical steering mechanisms completely.
The team tested this idea by playing the
game Mindflex. It reads the levels of
concentration from sensors in the headsets
Final Documentation
255
Mind Control Benchmark
and uses that to control the altitude of a
Conclusions
ball as well as whether it travels forward or
Although the mindflex game was a great
backwards along a track. It is comparable to
way to test out mind control activities, it was
how an EEG machine operates and senses
concluded that it is a very ineffective way
a patients vital signs. (Figure 31)
for a steering mechanism, since it was hard

to concentrate on performing simple tasks
(i.e. moving a ball up or down, forward or
Observations
backwards). It would be not feasible to apply
Being able to isolate your concentration
this method to complex tasks and steering
to particular tasks is not easy.
within a vehicle, even though it is possible
Only one task can be done at a time,
to train yourself to control and concentrate
hard to concentrate on two things at
properly.
once.
Once at extremes of high or low

concentration it is hard to move the
other way.
It is possible to train yourself to control

it properly based on the feedback of
concentration levels.
256
Final Documentation
Voice Command Steering Benchmark
would do smaller incremental operations
To evaluate how effective a user can steer
while turning left.
based on voice commands, the team set
voice. The team tested whether the drivers

trust themselves to perform this task and
the vehicle to provide correct information.
This helped to illuminate what types of
cues should be used in order to assist in a
transition or to allow more flexibility in the
driver position so that it is not necessary to
be looking out the windshield when driving.
(Figure 32: Blindfolded Test)
when
the
voice
left to indicate the driver should turn
experiment was performed in an empty

the passenger guided the driver with their
easier
commands used qualifiers like hard
up a blindfolded driving experiment. The

parking lot on Stanford campus in which
Became
the wheel a lot.
Interpretation of the commands was

very clear.
Conclusions
Several insightful conclusions were brought
up with this benchmarking idea. The voice
based steering functionality is not effective
since appropriate feedback and confirmation
cues were not provided. This leads to trust
issues and the driver feeling unsafe with
the situation. Although the voice commands
were clear and concise (left, right, hard
Observations
Hard to figure out if the steering wheel

was straightened out. It was also difficult
for the driver to drive in a straight path,
as they could not tell when to make the
slight adjustments that were necessary.
Difficult to tell whether we were even

moving (since moving at slow coasting
speeds).
It was very disorienting and made the

driver feel slightly off balance or uneasy.
Driver wanted some type of orientation

cues to understand why they were
being told to do certain tasks
Driver was very hesitant to go fast even

when told to just go straight.
Driver did not trust his or her own

reaction time to the commands.
When told to turn left, the drivers would

not turn all the way to left. Instead they
left, hard right), it was hard for the driver

to determine to what degree the command
should be taken. A hard right or left is very
ambiguous, especially when there are many
factors associated with driving including
speed, obstacles in the road, and the
sensitivity of steering for a particular car.
The number of levels in a steering feature/
mechanism is important to be able to make
small or large changes. The user input in
current mechanical steering design in cars is
a continuously varying input, as compared to
certain discrete degrees of input in the voice
commands based on the phrases used. The
ideal number of degrees required to steer
will fall somewhere in between. Another
important conclusion is that different drivers
would respond differently to such a system
based on past behavioral experiences. This
leads to the issues of past experiences
affecting driver psychology.
Final Documentation
257
Blindfolded Test
Haptic Command Steering Benchmark
Observations
The haptic command steering benchmark

was part of the same experiment and set up
The reaction time was longer in terms
like the voice command steering benchmark.
of noting the difference in command or
Since it was hard to determine what degree
when to stop doing one thing and start
of steering should be mapped to the voice
doing another.
commands (left, right, hard left, hard right),
another was not distinct enough.
haptic feedback, the team hypothesized,

should provide better commands in terms
The transition from one command to
Harder to perform when there were a
of interpretation. This benchmark only used
lot of tapping tasks happening one after
haptic command and not voice commands.
another
Tapping on the right shoulder meant to turn
Signals could be interpreted in different
right and tapping on the left shoulder meant
ways by different drivers, for instance the
to turn left. To indicate to drive straight,
stop tap signal could also be interpreted
both shoulders were tapped. To indicate
as a rapidly accelerate signal.
stop, both shoulders were pressed down
long periods of driving.
on at the same time. By tapping faster on

the shoulders with turning or going straight,
It was uncomfortable and annoying after
A lot of concentration and focus was
the driver knew to what degree it should be
needed to interpret/process the tapping
done. (Figure 33)
task and then to respond to the task.
258
Final Documentation
Haptic Command Benchmark
Conclusions
Segway benchmarking
Compared to the voice command steering
The team realized that in order for the
benchmark, although the reaction time

to interpret, process, and respond to the
tapping tasks was longer, the degree of
mapping of steering to the haptic feedback
seemed to be more effective than the voice
commands. Haptic and voice commands are
possible solutions steering mechanisms, but
more indicators to alert the driver prior to the
actual command would allow the driver to
anticipate the command (turning or driving
straight). Visual feedback still seems to be
the ultimate solution since the driver can
process the visual data faster and be able
to see the environment prior to it happening.
drivers to quickly take over the controls,

the control mechanism must be very easy
to learn. The team looked for an existing
example of a generally perceived intuitive
vehicle in order to isolate what was essential
to make a steering method intuitive.
A local entertainment center was contacted
and a time was scheduled to test Segways.
Segway uses unique steering method and
thus could be one of the future steering
directions. In order to go forward or
backward the driver needs to lean forward or
backward accordingly. Pulling the handle in
front of the driver to the left or right does the
steering left and right. The steering handle
serves also as a help to balance the driver
during the ride.(Figure 34)
Final Documentation
Observations
It is surprisingly easy to learn to drive

a Segway
Steering is very logical and similar to

body movements during walking
The small size of the vehicle and the

position of the driver (above the vehicle)
helps in precise navigation
Although the speed is not fast (20km/h

maximum) it feels like it is moving at
faster speeds
Conclusions
In order to make the driving experience
easy to learn, the team might try to use body
movements as one of the steering methods.
Being able to see the vehicle from above is
really useful in precise driving. If steering
is logical and similar to already learned
motoric body functions, it does not take long
to learn (it took about 2 minutes to learn to
drive Segways).
7.19.2 Motion Sickness Benchmark
The purpose of this benchmark was
to determine whether motion sickness
was an issue if no visuals of the outside
environment were given. The team initially
believed that focusing on a fixed visual while
having moving visuals in their peripherals
caused motion sickness. By eliminating the
moving visuals from the peripherals, motion
sickness could possibly be reduced. This
experiment involved covering the windows
of a car so the passenger in the back seat
could not see out of the windows towards
the front of the car. The passenger sat in the
back and read, while being driven around
campus. (Figure 35)
259
260
Final Documentation
Observations
Within 5 minutes of reading in the car,

passenger felt motion sick, even with
people who never experienced motion
sickness in the past
Good ventilation and lighting may

have been a factor the garbage
bags blocked off airflow to the where
the passenger was riding and lighting
produced moving shadows
May have been due to the extreme

driving that the tester was doing as well
Very difficult to read because you could

not anticipate turns or stopping
Conclusions
From our initial beliefs, motion sickness
Motion Sickness Benchmark
was still an issue even with the surrounding

environment visuals being blocked. Other
factors within our experimental set-up could
7.19.3 Human-Machine Transition
have induced the motion sickness problem.
Benchmarking
In this case, the remedy would probably be
The driver transition between autonomous
to have screens for everyone to see out from

the front into the horizon. Some sort of visual
and anticipatory feedback is necessary to
avoid motion sickness while doing work
in the car. Some options to explore would
be whether the complete view is required
or only the view from one side of the car
is sufficient. The problem with this is that
in metropolitan cities, having the view of
buildings zooming past the side could make
people more motion sick. Motion sickness is
definitely a problem and can be considered
an extension to the teams design vision
since the cabin space will be designed to
accommodate users doing desired activities.
and driving mode is an important aspect and

area of interest. To be able to determine the
steps that should be taken prior and during
the transition, the team observed two
different events that simulate the humanmachine interaction transition.
CNC Machine Operation Observation
The team observed fellow students using
CNC machines in the workshops to view a
scenario in which humans transfer control to
a machine. The team hoped to learn about
what kind of procedures occur before the
students transfer controls, what happens
when something goes wrong, and how they
transfer control back to themselves.
Final Documentation
261
Observations
Conclusions
Dry runs of the actual program were
The team observed that repeated visual
performed first before the final cut to
inspections and routines were performed
ensure the program worked properly.
initially to make sure that the program was
During the dry run the speeds could be
correct and doing what it is suppose to do.
controlled to be at 25%, 50% and 100%
In the situation of autonomous cars breaking
of the actual cutting speed for visual
down, having some kind of immediate
inspection.
service or real-time diagnostic analysis
Operator always had hands on the
can be performed on the spot. A major
stop button
finding was that dry runs with controlled
When something went wrong and did
speeds could be extrapolated as building
not know how to fix the problem, the
gradual trust with the autonomous system.
operator had to wait for a TA
It can either be different percent of manual
doubled
control and gradual shift to autonomous
checked that the correct drilling tool
thus building complete trust or it can also
was inserted.
be in completely autonomous mode with
The
operator
frequently
different levels of maximum speed. Trust in

the system and having self-confidence that
it is working properly is a main concern, but
gradually maintaining that trust is important.
Even though this experiment was focused
on the transition between human and
machine, it opened up many questions in
terms of developing trust. (Figure 36)
262
Final Documentation
Racing Game Player-to-Player Transition

In order to understand what it would feel like
to take over control of driving after being
in autonomous mode, the team decided to
experience this using a racing video game.
One player would drive the route in the
game, and then quickly give control over to
another player. (Figure 37)
Observations
Easier to switch if the other player was

observing what was going on
Racing Game Player to Player Transition
If the next player was not paying

attention, the chances of crashing
increased dramatically
Conclusions
From this experience, it seemed obvious
that a gradual shift of control is probably
better so that the drivers are aware of
the surrounding environments and of the
vehicles control actions. This way the driver
can align themselves or the controls in that
direction. An incremental rather than a direct
transition of controls would allow the driver
to get acclimated with what is happening
around them.
7.19.4 Confirmation Cue Benchmarking

Confirmation cue benchmarking is essential
in determining the psychological effects
of a driver in an autonomous car. Being
able to maximize the enjoyment of the
riding experience is important while also
keeping the user aware of the surrounding
environments and vehicles control actions.
Final Documentation
263
Light Indictor Experiment
anything, it made the passenger feel
The team wanted to test and understand
like the system wasnt working or wasnt
how much information would be needed for

the passenger in an autonomous car to ride
comfortably when all visual cues (i.e. seeing
out a window) are gone. The experiment
involved a passenger in the back seat of
the car in which they could not see out any
window or through the front of the car. A
console was lit up to indicate specific actions
of the car or surrounding environment prior
to the actions occurring. Two versions of this
quite sure what was going on.

Conclusions
Although the indicators seemed to be
effective at times, user acceptability will
depend a lot on the type and the amount
of information relayed, and how intrusive
it is when doing desired activities inside
the car during autonomous mode. One
possible further benchmarking to explore
experiment setup was tested (Figure 38):
would be to understand the different tasks
1. Console lit up to indicate what direction
finding a non-intrusive and effective way of
is being traveled (i.e. turning left, right,

driving straight)
2. Console lit up to indicate the speed at
which the car is travelling (i.e. slow to
being performed in the autonomous car and

giving notifications to the user. By having
anticipatory cues for out of the ordinary
situations, it will lead to reassurance that
everything is working fine.
fast)
3. Console lit up to indicate in advance
when something out of the ordinary
was going to happen (i.e. speed bump,
sharp turn, sudden braking)
Observations
Left/Right & Speed Indicators were not

really useful because the passenger
felt the motion. In this case, real time
indication is not useful.
Speed Bump, Sharp Turn and Braking

signs were effective because it gave
advance warning and was not real time.
One downside of this light system was

that the passenger had to be looking at
it to know, which made it irrelevant when
trying to do something else in the car
When the light wasnt indicating
Console in the car that indicated left,

straight, right (posts were replaced for
speed bump, sharp turn, braking indictors)
264
Final Documentation
Voice Indictor Experiment

The team not only tested how visual
confirmation cues affect how a passenger is
notified about the surrounding environments
and of the vehicles control actions, but
with the use of voice indicators as well.
The motivation of this was to see whether
giving only audio cues could provide the
passenger with enough information for
them to feel secure. The experiment setup
used the same setup from the light indicator
experiment, but the console did not project
anything. The rider in the front of the car
would say what is about to happen, for
example, stop approaching.
Observations

Voice is nice because it gives a more

precise description of what is going on.
Feels more secure and reassuring.
One concern is whether it is too intrusive

when music is being played or in
general.
Conclusions
Compared to the light indicator experiment,
the voice indicator seemed to be more
effective. User acceptability will still depend
on the type and the amount of information
relayed and how intrusive it is when doing
desired activities inside the car during
autonomous mode. The interface with
which the user interacts should maximize
the enjoyment of the riding experience but
still allow the user to be aware of whats
happening around them.
Final Documentation
265
Observations
7.19.5 Trust benchmarking

There
are
various
assistive
driving
technologies implemented in current high

tech cars that allow the driver to be in less
know that the ACC was working.
the team a better understanding of how and
to stop.
to decide what the desired following
vehicles in the future.
The ACC was tested within a Porsche

Panamera to show what feedback is given
distance was.
team was interested in finding out how

the driver engages the feature and how to
disengage or override the feature. It also
Override it at any time by pressing on

the accelerator or brake.
The visual of whether a car was in

front or not was helpful, as well as, the
currently to help ease the drivers concerns

about giving up control of the pedals. The
Easy to transition to using ACC by using

a lever to push in and a scroll wheel
why drivers will be able to trust autonomous
Adaptive Cruise Control (ACC)
No visual indicating how far away from

the car in front or that the car intended
control. By benchmarking and experiencing

existing driver assist technologies, it will give
There was no feedback letting the driver
following distance setting visual
Location on dashboard of the visual

was very convenient and only slightly
distracting
helped the team to experience what it feels

like to give up control in a way that the team
is not accustomed to. (Figure 39)
ACC Control
Panamera Dashboard
266
Final Documentation
Conclusions
the sounds are effective in alerting the driver,
It is important to have feedback to know that
it can get annoying. The interface that the
the system is working and what it plans on
driver interacts with is a crucial part in the
doing to make the driver feel comfortable
riding and driving experience, but being able
enough to relax and not be too vigilant by
to trust the systems autonomous functions
watching over the car. If the user sees what
are important too
the car sees, then trust is built.

Night Vision
Parking Assistant
The night vision assistance in the Audi A6 and
The parking assistant in the Porsche
A7 models is displayed on the dashboard of
Panamera has another feature that gave
the car. It is meant to help drivers at night be
the user more control and awareness while
able to easily identify obstacles in their path
parking. The display had yellow lines to
that should be avoided. When a pedestrian
show the cars projected path according to
is identified, they are highlighted with a red
the position of the steering wheel and has
outline to give the driver an alert that there
sensors to notify the driver how close the
is something in the road ahead.(Figure 41)
car is to an object. A beeping noise sounded

when the car was very close to hitting
Observations
something. (Figure 40: Parking assistant
display while backing up)
good
Observations
The use of infrared cameras is very

The image is clear, and the alerts and
identification works relatively well
This is very helpful in knowing if you are

turning the wheel properly, especially for
Conclusions
parallel parking.
Although the systems works properly and
The sensor/ beeping is slightly annoying.
effectively, it is a little distracting since the
It is very cautious and drivers realize that
display was bright at night and within your
even though the visual is red, they can
peripheral view which could draw your eyes
still move further.
away from the actual road. It is possible that
The display of the car and colors
it could be more helpful in nighttime/daytime
showing proximity to an object is very
driving in fog, snow, rain, or other hazardous
intuitive and easy to understand.
weather conditions.
Conclusions
The display of information on top of a realtime image of the actual environment was
very effective for parking and driving in
reverse during low speed activities. Although
Final Documentation
267
Night Vision Display
Parking Assistant display while backing up
Night VIsion
Lane Assistant
Conclusions
The team evaluated the lane assist feature
In order to trust the system, the user should
in the Audi A6 to determine how well it

functioned and what the experience was
be well aware how it works. Confirmation

cues of the system were really important,
with it engaged.
since the driver has to acknowledge that the
Observations
trust them. To really be able to do something
At first, the lane assist seemed a bit

scary, but became more comfortable
in time
The more the feature was used, the

more confidence the driver had in the
system
It felt strange that a simple lane change

on the highway without signal light was
so difficult when lane assist was on
It was unexpected when lane assistant

turned itself off
Was not instantly trusting left and right

warning light while changing lanes since
it is an unconscious, learned movement.
assistant systems were really on in order to

else besides driving, the cabin should really
be designed for it (space and tools).
268
Final Documentation
Night Vision Display
Camera
Audi A8 Equipment
Display
Camera
Final Documentation
269
Audi S8 2013 Benchmarking

The team got the chance to experience the
most expensive camera system (Audi Pre
Sense Plus) that Audi has to offer. This gave
a good insight of the current state of camera
technology in the cars and how it is used
for modeling the surroundings of the car.
(Figure 42)
Observations
Nowadays cameras can be quite useful

for observing the surroundings
Cameras make it possible for the

computer to model the surroundings
Corner cameras help to see what is

happening around the corner
If a person is walking to the front of the

moving car the human shape will turn
to red to warn the driver
The car tries to recognize human

shapes in the surroundings (it cannot
recognize animals yet)
Conclusions
In the future, computers will be able to
combine multiple sources of information
so that car will have better awareness of
its surroundings than a human driver. The
future car will most likely be able to predict
and calculate many different scenarios of
what will happen while driving.
270
Final Documentation
Golf Cart Prototype
7.20 Critical Prototypes

Critical prototypes are physical prototypes
uses in ME310 to identify critical functions
and
experiences.
Therefore
critical
prototypes are divided into CFP (Critical

Function Prototype) and CEP (Critical
Experience Prototype). CFP is a physical
prototype of a fundamental element of
design, which is required to ensure its
functionality and CEP is a physical prototype
of an experience of design, which is required
to ensure usability.
7.20.1 Critical Function Prototype Transition Golf Cart (Figure 43)

Driver comfort, safety and ease of transition
are important issues during transitioning
between autonomous driving and manual
driving for an autonomous car. It is important
to test the sequence of actions that would
take place and the kind of interface that the
drivers would interact with during the actual
transition.
The critical function being tested in this
prototype is the transition from autonomous
to manual. Situational awareness and being
in line with the actions being applied by
the autonomous controller are important
for a smooth and safe transition. That is
why the team implemented the concept of
Final Documentation
271
guided matching of controls (steering and
With Transition Golf Cart we tried to
throttle) during and after transition. This was
answer to following questions:
implemented by setting up a golf cart with
1. Do people feel safe and comfortable
a transition switch and an interface in front
making this transition using the
of the users which prompts them to mimic
prototype?
the actions of the autonomous controller.
2. Does the setup increase situational
The hypothesis being that such a guided
awareness for the driver?
matching interface will lead to increased
3. Are people good at matching?
situational awareness of the user leading to
4. Is the interface to intuitive?
a comfortable transition. Autonomous mode
5. What is the appropriate amount of
was faked using a passenger seat driver.
time required to make this transition
The actions of the autonomous controller
safely?
were displayed based on pre-recorded data
6. Is visual representation a good way
on the same track for driving. The details of
of conveying information to the driver
the setup can be found in appendices.
during driving?
Display
Golf Cart Prototype Driving
Switch
272
Final Documentation
The two types of transition sequences

that were tested using this prototype were
(Figure 45):
1) Gradual Transition
The user is given control of the pedals first
and a matching task is initiated. Once a
certain degree of accuracy is achieved in
matching the autonomous control actions,
the user is given the control of the steering
as well and a similar matching task is
initiated. After matching the steering, the
car goes into completely manual mode.
2) Direct Transition
The user is given control of both the control
inputs at the same time and a simultaneous
matching task is initiated. Once the user is
comfortable in matching, the interface goes
away and the car goes into completely
manual mode.
Gradual Transition
Transition
Match
Match
Initiated
Manual
Autonomous
Mode
Mode
Direct Transition
Transition
Initiated
Match
Autonomous
Mode
Gradual and Manual Transition Transfer Sequences
Manual
Mode
Final Documentation
273
Gradual
Direct
Pedal Matching
Simultaneous Matching
Steering Matching
Target
Regions
Step 1
Step 2
Observations
Users did not feel like driving and felt
unsafe during both the transition
sequences. People are more engaged in the
matching task than driving which is
supposed to be the primary task in
manual mode
In the gradual transition, once the users
are given the both the steering and pedals
in step 2, they neglect pedals completely
during steering matching
Matching error for the entire transition
sequence is lower for the direct transition.
Conclusions
The interactive interface was too distracting
during driving and it needs to blend with the
environment more easily. Visual
representation of information is too intrusive
during driving.
People are better at doing simultaneous

matching than matching the control inputs
one at a time
People are better at doing simultaneous
matching than matching the control inputs
one at a time
This might be because of the lower transition
Users felt safe and more comfortable in the
time in the direct sequence. The longer
direct transition
process involved in the gradual transition

delays complete situational awareness.
In the gradual transition, the users had a

tendency to grab the steering wheel as soon
as they switched to manual
Golf Cart Prototype Obeservations and Conclusions
The reason for this might be that having

a feeling of controlling the more sensitive
input, which is steering, makes the users
feel safe.
274
Final Documentation
The results were surprising because people
inputs from both the sources and that
were better at and more comfortable with
they would control the pedals based on
performing multiple tasks at the same time
feedback from the environment. However
than doing one task at a time. The reason
as observed before, the visual interface
for this was based on how information is
was too distracting and the pedal control
processed from various sources. As can be
was essentially an open loop. This was
seen in the flowcharts in Figure 46, there are
confirmed by the fact that the car came to
two sources of information involved the
a complete stop at times during steering
interface and the environment.
matching which meant that pedal control

was being completely ignored. On the other
In the gradual transition, as shown in the
hand, in the direct transition, the source
flowchart, it was expected that the users
of information for both the control inputs
would control the steering based on the
was the interface and the environment
Direct Transition - Expected
Direct Transition - Results
Interface
Environment
Gradual Transition - Expected

Interface
Transition Sequence Flowchart
Environment
Interface
Environment
Gradual Transition - Results

Interface
Environment
Final Documentation
275
Ideal Transition
7.20.2 Steering Mechanisms

Steering and pedals have traditionally been
Interface
Environment
the primary control inputs in automotive

design. When designing the cabin space
and transition interface for an autonomous
car of the future, it might be desirable to
explore other designs for control inputs that
will make driving more fun and be easier and
comfortable to transition to from any other
activity that is being performed in the car
space during autonomous mode. Having a
new control input might also open up new
options for cabin space design. This new
was completely ignored. But the users
control input needs to be intuitive to use
performance was still better because
and the users should be able to clearly and
there was just one source of information.
comfortably convey their intentions to the
Performance is not based on the number of
car controller.
tasks being performed at the same time but
To test this idea, the team hooked up
on the coherence between the sources of
different steering controllers to car gaming
information. As shown in the flowcharts, the
interfaces and users tried to steer the car
ideal solution would lie where the source of
with these non-conventional control inputs. A
information overlaps between the interface
pre-interview was conducted to understand
and environment. Thus in the ideal solution,
the driving habits and experiences of the
the interface blends in with environment
user. After the user tried out each of the
perfectly so that the goal of maintaining
prototypes, a post test survey was conducted
situational awareness while making the
which gathered information on the comfort
users feel safe and comfortable is achieved.
level of that particular method of steering

and its intuitiveness. The different types of
steering control inputs that were tested in
this prototype are:
Small hand controller
Steering with feet
Total joystick control
Joystick and control buttons mixed
Tilting feet to steer
276
Final Documentation
Based on the motivation for this CFP and
Steering with feet
some initial ideas, the team came up with
We wanted to test if steering with feet can be
certain requirements that need to be satisfied

by a good control input design. Metrics for
testing these requirements were identified
based on the car gaming interface setup for
this prototype.
a more intuitive alternative to conventional

steering methods. We purchased a steering
wheel game controller with pedals and set
up the game so that the turning left and right
is done by pressing left and right pedal and
acceleration and braking is done by pressing
Small hand controller

We used a normal Playstation Joystick tigh a
buttons on the steering wheel. Turning the

wheel doesnt effect the game whatsoever.
(Figure 48)
small steering stick which is used with thumb

movements. After initial testing with users,
Observations
we quickly realized that the tested steering
Turning was really difficult. Users didnt
method is not satisfying our needs and we
know how hard they should press the
quickly discarded it. Driving was way too
pedals.
imprecise using this method. (Figure 47)
Feet do not respond as quickly as hands

in such a dynamic situation.
The throttle button was binary and it

created some problems for the users.
Many of them wished to control the
speed better.
Having the steering wheel in front of the

users, without having any function was
Small Hand Controller
confusing. (The setting had a steering

wheel on the table where one button
is acceleration and turning was done
by feet by pressing left and right pedal)
Final Documentation
277
Steering with Feet
One person experienced fatigue in this

exercise.
One person replied that it would have

been better if she could see her feet.
People often made mistake and steered

to the wrong side
People who have never driven before,

said that they would like to have
someone controlling the gas and brake
for them. Whereas people who know
how to drive, would like to take more
responsibility and control the speed by
themselves.
Steering with feet is actually quite hard

to adapt to and has a steep learning
curve associated with it.
Steering with joystick

The team wanted to test if steering with a
joystick would be a more intuitive alternative
to conventional steering methods. (Figure
49) A conventional joystick was purchased
and set up so that the acceleration and
braking was done by pushing and pulling
the joystick and turning was done by pulling
the joystick left and right. In the second test,
turning was still done in the same manner,
but acceleration was changed to be the front
button on the joystick. (Figure - User-test
with joystick/Button on the joystick)
278
Final Documentation
Steering with Joystick
Observations
1. Control setting: four direction as
the speed as well.
Sometimes users didnt know if the
acceleration, move left/right, and brake
car was turning or accelerating. It was
(Figure 50)
hard to isolate only hand movement to

the side without moving it forward and
Joystick was not consistent in response.
backwards as well. Moving forward and
When it was pushed lightly, the vehicle
backwards is accelerating and braking.
in the game didnt react and if the
wants to brake he has to pull back on
too much. (Maybe there was a problem
the joystick, but the car in the game is
with this specific joystick or this game
still going forward for some time.
Like in the previous testing, it was
trouble for users. It was really hard to
difficult to keep car on the track going
keep the car going straight after turning.
forward after turning. Small directional
One person suggested to having a
adjustments are really hard.
joystick that moves left - right and not
It was confusing for the user when he
joystick was pressed hard, car moved
) That circumstance created a lot of
Hand hurts if the joystick is not placed
just to have left - right tilting motion
properly. Also it is painful to keep it
Steering in this way was easy to do with
pushed forward all the time. User
one hand
needed to keep it pressed in order to
It was hard to keep the same speed in
have constant speed.
turns. When user moves the joystick left

and right, he/ she accidentally disrupts
Some of the users brought the joystick

closer to their bodies
Final Documentation
279
1 .Control setting
2. Control setting: joystick left/right to control

turning, buttons to control gas & brake
Some users were performing better if
Accelerate
they have gas and brake separated
Right
from turning movements. One person

improved by 20 sec per lap after this
Left
adjustment. (Often they were still

pushing the joystick forward even if that
had no effect on the game.)
One guy suggested that it would be
Brake
better to have a speed limit and try

driving like that.
Braking button was not analog, so it

was always braking maximally instead
of slowing down the car.
Going backwards with a separated

button is confusing and it takes time to
realize what to do and where the button
for reverse is.
When users were able to see the whole

car, they almost didnt crash at all.
However, in top view it is hard to see
2 .Control setting
how much the car turns, because user

cant see the wheels in front of the car.
Accelerate
Right
Left
Brake
280
Final Documentation
Tilting Feet
The team wanted to simulated Segway-like
-steering for this prototype. In this tilting feet
setup the car was steered with tilting feet
to simulate the intuitiveness of a Segway.
(Figure 51)
Observations
Accelerating same time with tilting

forward and steering proved to be
almost impossible
Joystick position on the floor was

almost impossible to put in a way that
is comfortable to steer with
It took too much effort to tilt the joystick

to its full extent with just feet.
Tilting Feet Controller
Final Documentation
281
Another insight is that listed requirements
from CFP
should not be applied in our context since
The assumed purpose of this prototype
we realized that the intuitiveness is not really
proved to be quite irrelevant. It was
an issue. These steering methods can also
discovered that none of the non-conventional
be tested in perspective of fun in the future.
steering options are that precise and all of

them have a steep learning curve associated
Opportunities
with them. However, the most interesting
Steering, accelerating, and braking is
insight from this prototype was that manual
entertainment that cannot be done the
driving is redefined in the future with cars
same way in anywhere else. Simulators
being semi-autonomous and crash proof. In
and games are always missing some
such a situation, it might even be desirable
parts of the holistic experience.
to have a non-precise controller that gives
The intuitiveness is not the issue, since
the user an illusion of precise control while
driving safely will be enhanced with
making driving more fun.
active safety systems that autonomous

car and its infrastructure can provide.
Almost all of the cars in the future will be
It might also be safe to assume that
part of the autonomous car system since if
passive safety systems (seat-belts,
one person is allowed to drive outside the
airbags, etc.) can be removed and the
system and do unpredictable moves, the
freedom of cabin design is increased.
rest of the drivers will be in danger. This is

why intuitiveness is not really an issue from
the perspective of learning. After testing this
prototype we will approach steering from
another perspective: the fun perspective.
Autonomous technologies make totally
safe driving possible and driving in a semiautonomous mode lets the driver to push the
boundaries knowing there is no dangers in
driving because the car takes care of safety.
However the driver should not feel like the
car is applying a lot of corrective action even
in manual mode. That needs to be avoided
since the team aims at maintaining the
pleasure of driving through the proposed
design.
282
Final Documentation
7.20.3 Critical Experience Prototype -
Reconfigurable Workspace
Workspace prototyping is an important
desired activities
when the user wants to take control
be in autonomous mode a majority of the
ASAP
time, being about to do the drivers desired

vehicle is ideal. The driver not only needs
to be able to do their activities in the space,
but it needs to be a comfortable and more
The team wanted to explore the experience
with this setup. It has to be fast enough

for a quick transition but slow enough to
a mock cabin space with movable chair to
avoid discomfort.
be good to have the steering apparatus
quick prototype of the envisioned cabin
attached to the moving chair so that you
space with an interactive windshield display
can control the car from any location
and a moving chair for repositioning the user
and orientation within the car
the cabin space in an intuitive way for
prototype is shown in Figure 52. The details
Observations
The setup of the moving chair

transitioning to the cabin activity space
table was an interesting feeling, not able
to do that in current cars
Position of person when moving forward

or backwards in transition is awkward
especially with the feet dangling
It would be desirable to have an interface

that displays various configurations of
activities like working and relaxing. The
been included in.
One interesting concept that came out

of testing this prototype was that it might
To test this experience, the team made a
about how this prototype was tested have
It was observed that the chair speed

affects the level of comfort users have
of a car. The environment was setup to have
has been adapted to perform some of the
Users noted that the experience was

not like being confined in a typical car
of having a moving chair in the cabin space
during autonomous mode to a location that
It is desirable to have more room and

space to do various activities
intimate personalized experience.
transition from driving to leisure mode.
Timing may be a problem when trying to

switch in transition suddenly, especially
aspect of the design. Since the driver will
activities within the cabin space of the
Comfortable with space to be able to
the user to interact with.
Besides having a switch, there can

be other ways of conveying the users
intention to the controller like voice
based and gesture based commands.
Final Documentation
283
Interactive
Windshield
Movable
Chair
Cabin Activity
Space
Reconfigurable Workspace
284
Final Documentation
7.20.4 Critical Experience Prototype -
Results
Mobile Workspace
The team wanted to explore the experience

of how it actually feels to work in a moving
vehicle and whether it is even possible to
do that comfortably. Our assumption is that
some of the time used in autonomous car will
be used for working. If working is possible in
a simple setup, it will certainly be possible
Nobody was motion sick, although it

was not possible to look outside.
It was easy to concentrate in the van,

because of the lack of distraction
It was hard to write on the whiteboard

while moving
Some of the users were completely

relaxed during driving, but others were
in a space designed for that.
constantly aware of the environment
Cabin space could also be used for
small window between front and back
entertainment. The aim of this prototype was

to create a situation where a passenger in
the future car can do some physical activity
and maybe have fun during the drive. This
was implemented by trying to play with a
Nintendo Wii, because it requires physical
movement. Different physical activities were
simulated in the car by having different video
and tried to follow the driver through the

of the van. It might be possible that the
confidence comes with familiar driver.
It is possible to work in the car while moving.
The team was not so sure after car tests
made earlier, but when tested out with this
van that was moving slowly, it was really
games on board.
comfortable to work there.
Test Setup
reasons behind motion sickness and what
The team rented a van in order to test the

possibility of having an office and to work
in a moving vehicle. Within the van two
different scenarios were created (Figure 53):
Office scenario with a table, chairs and a

whiteboard in the back of the van where
a meeting was conducted while the car
was moving slowly to simulate riding in
an autonomous vehicle.
Entertainment scenario included having

a gaming console (Nintendo Wii), a TV
and a fatboy bag inside the van.
The team needs to explore more the

visual or other cues need to be provided to
prevent it.
Final Documentation
Mobile workspace - Office Scenario (top), Entertainment Scenario (bottom)
285
286
Final Documentation
follow to make the transition easier. This
As described in Section 4.5.6, one
the steering and pedals being displayed
of our main CFPs was the one that

tested out the transition sequence by
simulating autonomous mode on a golf
cart and prompting the user to mimic
the actions of the autonomous controller
for situational awareness. This section
lists
the
design
specifications
and
detailed description of this prototype.
plan required real time information about

to the user. The following sections give
details about the electronics and the coding
strategy used. The entire code can be found
in Appendix 7.6.2.
Mechanical Setup
As shown in Figure 54, the mechanical
design for this prototype consisted of
designing a means of indicating the desire to
7.21.1 Transition sequence prototype
switch to manual from autonomous by using

a switch. This switch needs to be placed in
The main objective of this prototype was to
a comfortable and accessible position for
test the comfort level and intuitiveness of a
the user. In this prototype this switch was
gradual
from
mounted on the cover that was designed to
autonomous to manual mode. As described
protect all the electronic circuits that were
in Section 4.8.1, both the transition
mounted on the steering wheel of the golf
sequences consisted of real time data being
cart. The proposed design for this prototype
displayed to the users and along with pre-
also required an interface right in front of the
recorded data that they were supposed to
users while driving to prompt them to mimic
or
Transition Switch
direct
transition
Display Construction
Final Documentation
287
the autonomous controller actions. A 7-inch

tablet was used for this purpose as shown in
Figure 55. To create a mount for this tablet,
the team bent a sheet of acrylic to make it
a holder for the tablet and added mounting
clips on the same for securing it to the golf
cart windshield right in front of the user. at
the same time. The reason behind it was
to learn about peoples working habits in a
relaxed atmosphere.
The way these two designed components
were mounted on the golf cart has been
shown in Figure 56. It should be noted
that the laptop was just for real time serial
communication with the sensors and kept
on the passenger side of the car in a way
that was not interfering with any of the users
actions during testing.
Golf Cart Construction
288
Final Documentation
As shown in Figure 57, below the electronics

were mounted right on the steering wheel
beneath the designed component for
protection from damage during testing. This
also turned out to be a useful shield from
rains during testing. The mechanical design
allows the flexibility of flipping open the
cover and working on the electronics if the
designer wants to change something on the
breadboard or check for lose connections.
Switch
Switch Contrustion
Final Documentation
289
Description of the electronics

Arduino duemilanove microcontroller was
used for receiving data from the sensors
3V3 5V
Vin
and communicating with the laptop.
RST
D13
A software platform called iDisplay on
AREF
D12
Android was used to extend the screen
D11
Arduino
of the laptop to the tablet for displaying

the interface that was coded in Visual C#
Arduino board. There were two sensors

used for getting the steering and pedal
data respectively. There is an internal
potentiometer mounted on the throttle pedal
of the golf cart to give the throttle signal to
D9
D8
D7
and which responded to the data received

through serial communication with the
D10
D6
A0
D5
A1
D4
A2
D3
A3
D2
A4
D1
D0
A5
GND
its internal controller. There was an external

tap created on the voltage reading of this
potentiometer and the value was fed to the
Figure 58: Arduino Board
Arduino ADC. It should be noted that only

one pedal (throttle) was used for this entire
experiment and the brake pedal was not
used. The analog input sample circuit has
Acceleration of the car itself and sudden
been shown in Figure 58.
bumps will make the net acceleration

vector jump around 65
The steering angle was detected using a
Rotational acceleration while turning the
maximum acceleration vector detection
steering wheel will also cause problems
algorithm applied to the data received

from the accelerometer ADXL345. The
However this was a basic prototype to show
connections for this module have been
proof-of-concept and a rough steering angle
shown in Figure 42 above. When the system
calculation is fine for this purpose. It was
is reset, it takes 5 seconds and notes down
actually working pretty well considering that
the stationary initial net vector (which is
the golf cart was not accelerating too much
mainly just gravity). Once the algorithm
on the track and there were no bumps. Also
starts, every time it calculates the angle of
the accelerometer was mounted, as close to
the current net vector with this initial vector
the center of the steering wheel as possible,
and converts it to the steering angle. There
so there was no or little effect of rotational
are two ways in which we will get inaccurate
acceleration on the Cartesian acceleration
readings from this method:
vector that was being measured by the unit.
290
Final Documentation
Description of the code
to manual command is received. When the
Figure 59 outlines the flowchart used for the
user flips the switch, the display comes live

and the transition sequence is initiated.
code. In the recording phase, the car is ran
The accuracy of matching is checked by
once on the pre-decided track with steering
incrementing a loop counter inside the code
and pedal data is recorded as a reference.
and showing the score in a progress bar on
This recorded data is treated as the ideal
the right in the interface. If the user goes
actions that the autonomous controller will
outside the desired region the counter is
apply and which the user is expected to
reset to zero. Once the user achieves the
follow while trying to retrace the same path
desired level of accuracy the interface
and transitioning into manual mode.
switches off and the car goes into completely
The code is continuously communicating
manual mode.
with the microcontroller to get real time

data, but it is not displayed until the switch
Arduino Flowchart
Stop
Select Mode
Write data
to file
Recording
Outside
threshold
Accuracy
achieved
Remove
Display
Loop
counter++
within
threshold
Wait for
switch
Switch to
manual
recieved
Display recorded
and real time data
Final Documentation
7.21.2 Different Steering Controls

The another CFP made was to find out the
most intuitive steering method out of many.
This section lists the design specifications
and detailed description of this prototype.
Joysteer prototype
The reasoning behind this prototype
was to discover the most intuitive way of
291
steering since we assumed that the drivers

will drive much worse in 2035 because of
autonomous features. The prototype was
realized by building a test setup of video
game and different controllers to steer a
car game. The user testers were given a
one controller and one track to finish. The
following section will illustrate what kind of
test setup was built. (Figure 60; 61)
Basic Setup
Basic Setup - TV, Audi A4 car chair, Playstation 3, Colin Mcrae Dirt for PS3
292
Final Documentation
Final Documentation
293
Test Setup:
Small joystick control (Figure 62)
Large joystick control: Saitek Aviator PC
(Figure 63)
Thrustmaster Experience Racing Wheel
with Pedals (Figure 64)
Sponge (Figure 65)
Duct Tape (Figure 66)
Reboard
Racing Wheel
Small Joystick Control
Sponge
Large Joystick Control
Duct Tape
294
Joystick as a tilting controller setup
Sponge
Reboard
Joystick as tilting controler setup
Final Documentation
Final Documentation
295
Joystick as a tilting controller

For using joystick as a tilting controller a
sponge was cut and paired with a large
joystick. Tester needs to put feet on top
of reboard and use feet tilting to steer and
manage speed.
(Figure 66: Joystick as tilting controler setup)
Controllers
Gas and Brake
Steering
Small Joystick
Buttons
Analog control
Large Joystick
Analog up & down
Analog control
Large Joystick
Buttons
Analog control
Pedals
Buttons
Left & Right pedal control
Feet tilting
Analog up & down
Analog control
296
Fall quarter was primarily used to gain more

knowledge about the task we were given.
We used the time to benchmark related
industries and tested prototypes mainly
based on our assumptions.During fall
quarter we tested different ways of steering
and discovered that the steering wheel was
Final Documentation
7.22 Needfinding
7.22.1 Trip to Germany
After Christmas break, the team got an
opportunity to visit Germany and had the
first meet with corporate partners from
Audi, which was the main purpose of this
still the best way to steer.
trip. Since we already traveled to Munich,
In the beginning of the winter period we have
and BMW Welt, as well as Audi Museum.
already had understanding of the problem,

future assumptions, user and the technology
that might be available in the future. We had
a meeting with our corporate partners in
January and by then we had entire picture
of what was expected from us to deliver and
we also decided to drop in BMW Museum

In addition to previous, we also got an
opportunity to visit University of Munich
and see their automobile laboratory and
jet engine laboratory. The trip itself was
genuinely hectic but also highly educational.
Travel plan for our travel can be found in
what might be our final goal.
Appendix.
Also, we tested different activities in the car
List of our targets in chronological order:
and realized that the size of the vehicle was

the only boundary we would face when it
came to utilizing the time in the car. Space
in the car was limited and making the interior
of the car suited for different activities was
1. BMW Welt
2. BMW Museum
3. Audi Museum (Figure 68)
4. 1st Meeting with Audi contacts Lorenz
Bohrer and Tilo Koch
our biggest challenge.
5. Audi Factory Tour
In addition, we tested the transition between
Bohrer and Ulrich Mueller
autonomous and manual driving. In order

for the driver to be confident in cars ability
to steer itself and to know that the transition
of control has been made, there has to be a
whole sequence of clues and reassurances.
Because of its complexity, after winter
period, transition from autonomous and
manual driving still remained a problem that
we needed to research more.
6. 2nd Meeting with Audi contacts Lorenz

7. University of Munich, Vehicle Laboratory
8. University of Munich, Jet Engine
Laboratory
Key notes
OLED technology is going to spread in

car interior design
We should narrow down our scope
Audis formula for successful products

is
sophistication,
sportiness
and
progressiveness
1/3 of people get simulation sickness
Final Documentation
297
(same as motion sickness) because
of the mismatch with senses and
The trip itself taught the team about the
predictions of brains. -> Interesting
long history of German car industry and
statistics for us.
its significant focus on detailed design as
Audi is researching SbW (steer by wire)
well as high end engineering. Audi itself is
technology and it will most likely be
really passionate about finalized products
introduced in future cars
that are high quality and well designed, they

were constantly emphasizing these matters
More notes and some explanatory pictures
greatly and wanted to be sure that tangible
can be found in Appendix.
prototype we provide for EXPE will meet

their requests.
298
Final Documentation
7.22.2 Geneva Trip
Smart for two, Aston Martin Cygnet...)
Geneva car show (Figure 69), as one of the
or they offer slightly bigger versions for

four passengers.(Fiat 500, Opel Adam,
biggest car shows in the world, annually
Citroen DS1...), but also there are many
attracts the most famous car manufacturers
vehicles between already established
to present brand new car models and
classes, like small SUVs, Expensive
concepts. Many of them have their world
sport cars hatchbacks (Ferrari FF, Aston
premiere there. It was a unique opportunity
Martin Rapide...) or even cars between
for us to see what the current and future
motorcycles and passenger vehicles.
trends in car industry were, to speak with

professionals from the field, to observe
design of the cars, and to locate visual
(Renault Twizy, Toyota Iroad...)
Current trend in the visual appearance
elements that make a car looks futuristic.
of the cars is to have lines which are
Observations
manner. Lines are not parallel to the
ground any more, but they flow freely
Electrical cars are a reality. Every major

car manufacturer already provides an
electric vehicle.
braking the surfaces in not so logical
Car manufacturers are constantly

expanding the number of models they
offer. Many manufacturers already
have small vehicles (size of Toyota IQ,
Geneva Car Show
on the surface of the car.
LED lights and glowing lines in

headlights are common
Even small, compact vehicles tend to

look aggressive
Final Documentation
299
Results
Combination of colored LED lights

(usually
light
blue),
transparent
were not known or wrongly interpreted
materials and playful mixture of different

lines and patterns on the chassis of the
by 12 testers.
Indicators like engine oil temperature,
car make the car look more dynamic
ABS, cruise control, airbag, check
and futuristic
engine light, battery level, were not
There are not many changes in the car
always needed.
interiors. Mainly, changes are related
Halves of the logos on questionnaire 1
Motor temperature, oil temperature, and
with introducing new technology, like
battery life were considered the least
touch screens.
important indicators in questionnaire 2
It is possible that in the future there
by 16 testers.
will be more vehicles for one or two

persons.
Compared with dashboards in 20 years ago,

7.22.3 Dashboard Questionnaire
With dashboard questionnaire we wanted
to explore how aware of different indicators
that are on a dashboard the users actually
are. Rationale behind this was to see what
information is actually valuable for driver to
have when he is using a car. We had two
different kinds of questionnaires (Figure 70
and Figure 71). In 1 questionnaire we asked
drivers to identify what each logo was for
on on the dashboard. In questionnaire 2 we
asked drivers to arrange the meters in order
based on their opinions from the most useful
to the least useful.
1.
3.
Dashboad
Questionaires
2.
4.
5.
6.
dashboard is more simple, showing less

information. We may get rid of dashboard
and only show the most important indicators
on augmented reality windshield. Drivers
can choose customized information that
they want to see. Some information like fuel
condition and turning lights is only shown
when needed. This way we get more cabin
space. This area may be utilized for working
space.
300
Final Documentation
Changing
Manual / Autonomous
Brake buttons
Acceleration
Forward, Reverse
and U-turn mode
Link and Go - Car Concept
Rotatable Chair
Final Documentation
7.23 Benchmarking
7.23.1 Akka Car Concept
301
Steering
but it doesnt spin when the car is in
Among other cars at the Geneva car show,

there was one car which focuses on the
same issue as we do. It was a working
autonomous mode.
handles behind the wheel and braking is
Akka Technologies in collaboration with

The concept is called Link and Go (Figure 72)
and it offers both manual and autonomous
driving experience.
Observations
Outlook of the car:
Car doesnt look sporty or aggressive.
done by pressing buttons with thumbs.
in order to drive.
When driving, the seat can be locked.
Drivers seat is really not ergonomically

designed for comfortable driving, but for
multifunctionality and changing of the
modes.
It is slightly taller than normal cars. It

of freedom.
The whole car is larger and has more
Performance
Back seat looks like a piece of furniture

and the whole car looks more room like
on the roof of the car.
Front seats are attached to each other

and are rotatable 360 degrees. Driver
can turn the front seat entirely and
socialize with people in the back.
During the rotation of the seats, there

is not enough place for legs.
Changing between modes is done

before the drive. There is no option
for changing modes on the go. There
is a display attached to the front seats
where user can set up the mode.
The car is designed for city use. It has

a maximum speed of 50 km/h when in
autonomous mode and 120 when driven
Mode changing
The car is a working prototype. It has

sensors in the front and a big sensor
space in the interior.
When in driving mode, driver needs to

manually pull the steering wheel closer,
for convenience and not for pleasure.

provides more comfort and the sense
Steering wheel moves closer to the

dashboard when not in use.
It is mainly a functional vehicle which is
Acceleration and brake pedals are

removed. Acceleration is done by pulling
prototype of an autonomous car, done by

other companies that provide software.
Steering wheel is always visible,
manually.
Car is meant for the year 2025 because

of the legal issues.
302
7.24 Dark Horse Prototypes

Dark Horse prototype is the term we use to
describe a prototype which reflects our most
radical and risky ideas. It presents a leap of
fate, something that might work and maybe
can be implemented in the final concept or
as the concept itself. In this exercise we
were encouraged to think outside of the box
and this was our last chance to experiment
and to test unconventional ideas before we
focus on something which will eventually be
our final concept.
Because of the lack of time caused by
our trip to Germany, we had limited time
to develop highly functional prototype,
basically we only had only one more
sophisticated prototype. In addition, we
decided to make a few of rapid prototypes
of our ideas and to present those to users
for some feedback and further inspiration.
In the following paragraphs we will explain
these prototypes briefly.
Final Documentation
7.24.1 Dark Horse Prototype Reconfiguro

The dark horse prototype tested whether the
users of autonomous vehicles would value
the concept of a reconfigurable interior
space. A reconfigurable space would allow
the users to pick and place the objects and
furniture that would be an appropriate fit
for the activity they would do in the vehicle.
This prototype also determined what level
of control the users would want to have over
the reconfigurable space.
Setup
To create a realistic experience for this
prototype, it was pertinent for the users
to feel what it would be like to reconfigure
the interior space from the outside and
then perform various activities within the
space. The team rented a U-Haul moving
van to replicate the autonomous driving
experience, as show in Figure 73. The
can already had a partition between the
front and backspaces so the user had no
visuals of the front and the driver. The
objects that were available for the user
Final Documentation
303
Reconfiguro Setup
to configure the vans back interior space
prior to getting into the van to configure the
included two chairs, an inflatable bed with
space to give them a realistic experience
pillows, a low-rise nightstand, a high-rise
of what it would feel like to control the
desk, a TV projector, a desk lamp, a deck
cabin space remotely from anywhere. The
of cards, drinks, and magazines. The TV
vehicle was also equipped with speakers
projector was used in conjunction with
not only to play the audio from the TV
a laptop computer to display TV shows
shows and music, but to amplify a text-to-
on the vans partition screen (if the TV
speech application to imitate a robotic car
projector was selected in the interface) or
voice. The text-to-speech voice was used
a front-view driving visual (if the projector
to communicate with user of their travel
was not selected), as shown in Figure 74.
updates, ask the user specific questions, and
The user interacted with a tablet interface
to respond to any requests from the user.
304
Final Documentation
Control Interfaces
The second interface was a parameter-
Three different interfaces were tested within
based application (shown in Figure 76)
this experience to determine what level of
in which the user would answer specific
control the users would want to have over
questions pertaining to the parameters of
the reconfigure space. The first interface
the activity or drive. The user would answer
was a Java drag and drop application
questions such as how many passengers
(shown in Figure 75), which the user could
will be riding, how long will you be riding,
have full control over where and in what
and Set car for: (relax, sleeping, socializing,
orientation the objects, could be placed in
etc.). Unlike the drag and drop interface,
the interior. The drag and drop application
this parameter-based application gives less
was designed to portray the objects in a
control to the user in the type of configure
proportionally realistic manner relative to
that is possible within the interior space.
the size of the vans interior space which
The slides for this interface can be seen in
allowed the user a better perception of what
APPENDIXSection 9.8.3.
could really fit in such a space.

The code for the java application can be
found in Section 9.8.2 APPENDIX.
Control Interface
Final Documentation
305
Control Interface - Parameter Based
The third interface that was tested was
The drag and drop and parameter-based
based on the concept of a lifestyle sync
interfaces were presented on a small tablet
application in which the autonomous vehicle
device to give the user a realistic experience
knows the users schedule for the day and
of what it would be like to configure the
rearranges the space based on what kind
vehicles space remotely from anywhere.
of things the user will be doing throughout
The lifestyle syncing interface was given
the day. The vehicle also talked to the user
to the user as a survey to determine what
and intuitively knows if the user needed
their schedule would be, and then the space
something specific. For example, if the
was configured for those activities.
user just went to the get a Starbucks coffee

but there is no cup holder in the previous
configuration. The vehicle will know and
prompt the user if he or she needs a cup
holder. With lifestyle syncing, the vehicle
would train and learn from what the user
needs, requests, and actually does to be
able to configure the vehicle in the most
optimal way.
306
Final Documentation
Testing Procedure
was very similar to flow 1 and 2 except
The testing was performed on campus with
that the user did not reconfigure the space.
different stops locations around campus:
Another additional feature within flow 3 was
1. Starbucks to pick up a cup of coffee
the ability to communicate with the van by
2. Work to pick some books
asking specific questions or requests and
3. Trip to pick up a friend
the car would respond.
Figure 77 portrays the different experience
The behind the scenes action of how the
flows for each of the interface interactions
whole experience flow worked was very
and demonstrates the behind the scenes
important. Two of the team members
procedure to reconfigure the space.
operated the van environment; one person
Experience flow 1 and 2 correspond to the
was driving while the other person was
drag/drop and parameters based interface.
operating all the internal features. The other
The user was given the story about that they
member followed the van on the routes with
just received an autonomous car and they
all furniture objects that were not in the
were going on three different stops around
cabin space. Once the van made a stop for
town (stops were performed on campus).
the user and the user went off to perform
The user dragged/dropped where they
their tasks, the entire team reconfigured the
wanted the furniture to be placed within
cabin, swapping objects from car to van.
the cabin space or answered questions

pertaining to the parameters of the activity
and drive. Once the space was configured
accordingly, the user hopped in the back
and proceeded to the first stop, a 5-minute
journey. The user picked up a snack or
coffee and right before heading back to the
van, they reconfigured the space to their
needs or desires. The next stop was a
20-minute journey and the user performed
similar tasks as during the first stop. The
last stop consisted of picking up a friend
and socializing.
Experience flow 3 corresponded to the
lifesync application. The van knew the
users schedule and preferences and made
an educated guess at how they wanted the
cabin space to be configured. The flow 3
Final Documentation
307
Experience
Flow 1 and 2
Experience
Flow 3
Behind the
scenes
User comes
Introduction story
marrated. Schedule
given.
User comes
Introduction story
marrated. Schedule
given.
Two team members

ready near the van
User selects
configuration from
the tablet interface
User answers
preference survey
User reaches car

and starts with short
5 min journey to
grocery store
User reaches stop 1

and goes shopping.
Uses tablet interface
to re-change
configuration.
Long Journey to
work started based
on the configuration
chosen.
After work, car

reconfigured for two
people and socializing with a friend.
User reaches car

and starts with short
5 min journey to
grocery store
Users go shopping
and return to find
space reconfigured
for them.
Long Journey to
work started. Car
prompts user for
things they would
like to change.
Changes are made
by making a quick
stop.
After work, car

reconfigured for two
people and socializing with a friend.
Car still prompts for
comfort with current
configuration and
desired changes.
Cabin space
reconfigured based
on what user wants
and team is in
driving position.
Two members of
theam in the van.
One member
follows in another
car with spare
furniture.
Cabin space is
reconfigured based
on what user wants
and team is in
driving position
One team member

follows in van and
helps in changes
on the go.
308
Final Documentation
Results
There were a six user tests that interacted

with different interfaces: one test with the
It was hard to gauge how cluttered it will

actually be inside the van
Initially the users tended to place as
drag and drop application, one test with
many objects as they could in the cabin
the parameters based interface, and four
space
tests with the lifesync application. More
When they were in the car they spent
user tests were performed with the lifesync
a significant portion of their time
because more valuable insights were
discussing or thinking about what the
collected during the first initial test. Figure
next configuration will be.
78 displays some of the users cabin space

configuration.
Common observation for conditions for
Parameter based interface:
It was easier than the drag and drop.
The configurations created for the users
experience flow 1 and 2
were acceptable and they did not need
that much level of control
Having to configure before getting into

the car interrupts their daily activities
and in general, it is more effort on the
LifeSync Interface:
users side
It made transitioning from one place to

another in your daily schedule seamless
Drag and Drop Interface
The users did not mind making minor
It took the users a long time to decide
adjustments to the configuration while
on the configuration
they were travelling to better suit their

needs.
User Tests Configuration (bottom and right)
Final Documentation
309
The users liked the fact that the car
Facing each other was most conducive
anticipated their needs and gave
for work meetings and socializing.
suggestions
the
Having them sit right next to each
configuration that the user might desire
other can be useful for multimedia or
for
changes
to
for romancing.
General Experience Observations:
The person that gets picked up views
While relaxing, the users did not always
the other person who was riding the
want to be lying down completely, would
autonomous car before as the owner. -
prefer something like a reclining chair.
(In one instance, when the car asked a
Users reposition themselves a lot in the
question to the users and they replied
cabin space.
with different opinions then the person
Just the front projection of driving is not
who owns the car gets priority in group
enough to comfort the users and make
dynamic. In the second instance, the
them feel safe
friend did not really interact with the
Maintaining
visuals
not
just
for
car as much as the owner, as they are
situational awareness and safety but
under the assumption that they owner
also for pleasure is important
knows better how to interact with their
People get used to voice commands
car.)
while interacting with the car very
Rider became embarrassed when the
quickly and if it stops talking to them
car asked something that revealed
then they feel that something is wrong.
personal information.
There is an emotional connection that
When a user was riding sideways, their
is developed with the car when the car
body was turned slightly towards the
talks to the users.
front.
310
Final Documentation
If the car did not establish rapport early
from the darkhorse prototype, the team
into the ride they felt less comfortable in
wanted to focus on the aspects of mobility
speaking with the car. People got really
and flexibility within a smart cabin space
comfortable very quickly after this.
design, and concluded that within the cabin,
Since it does not feel like a car anymore,
the chair is an important factor that affects
you need more accessories like trash
this design space. One possible solution
cans, outlets, wifi.
was to design a chair that senses whether
While talking to the car, the users want
the user is falling asleep or if they wanted to
to humanize it as much as possible.
rotate and sit in a different orientation and
Things kept falling off the table.
it just does that without the user inputting
Having just a fixed cup holder makes it
any specific commands. So is it possible
constrained a lot.
to sense a users intentions and adapt the
Car not interrupting the conversation of
chairs position based on the users body
a group or knowing when to interrupt
movements while not intruding in the users
the conversation. It is essentially like
activities?
another human that must follow all the

social etiquettes.
Car
user
confidentiality
what
information is revealed in front of the

users friends and what is not.
Changing configuration when the car

has stopped was better. So there is a
need to change configuration safely
when the car is moving.
The users forgot that they were in

the back of the van and when they
saw other cars in the projection, they
suddenly realized that they were also
7.24.2 Dark Horse Prototype - SbW

Imitation
Night before leaving to Germany, we
wanted to test with a quick prototype if it is
possible to steer in non-ordinary positions,
for example sideways or back towards front.
(Figure 79)This prototype was based on our
assumption that future autopiloted cars will
be aware of the surroundings so that optimal
driving position is not necessarily required
for safe driving.
in a car and the projection was in fact
To see if driving is actually possible in
distracting to the main activity in which
various positions, we could maybe place
they were engrossed.
the steering elsewhere in the car and this

way acquire more freedom in cabin design.
Conclusion
We wanted to also test how does it feel
People value the benefits of being able
to have contradiction between expected
to optimize the cabin space for different
vehicle movement and the actual vehicle
activities but they would like to do so in
movement.
the least amount of effort and intrusion in

their daily lives. From the insights collected
After visiting Audi, they kindly suggested us

to focus on some other factors inside the
Final Documentation
311
cabin. They also told us that they already
Setup
have an own laboratory focusing on this
Skype call between iPhone and iPad ->
so called Steer-by-Wire -technology. The
quick analog setup for screen/camera
reason they are studying this is mostly
solution
because it allows the extra freedom in
iPad was also used as a steering wheel
dashboard are and also because in future
One person was pushing and steering
electric cars are mostly done with this
the vehicle according the drivers
technology. This is because it reduces car
gestures
weight and the car is already eclectic and
Before iterating setup any further we
would not function at all without electricity?
quickly tested steering sideways and
This sort of digital technology replaced
back towards front
analog speedometer in late 1970s and it

is going to eventually replace the analog
steering.
Iphone with camera,
giving a vision forward
Ipad as a wheel and screen
312
Final Documentation
Observations
Vehicle can be steered in non-ordinary

positions in slow velocities without
difficulties to adapt
Minor crashes were caused by rough

setup since the external person steering
everyone would like

to sleep in the car if it
would be possible
cannot react right after the driver has
2. Driver sleeps in a relatively uncomfortable
given the input
position so that he/she can enjoy a nap, and
There is also a little delay between
the car is able to wake the driver up in a
the phone and iPad which also makes
relatively short time.
steering a little difficult

Sleeping in first generation autonomous
cars, which are going to be introduced within
We can assume that the future electric
10 year, sleeping is not most likely going to
cars are going to use this SbW-technology
be possible since reorientation from sleep
since it already under a lot of research at
takes too much time (about 30 seconds)
Audi. The delay issue can also be solved
and and early autonomous cars have to
with more sophisticated technology. To test
be able to give the controls for the driver
functionality of steering through screens in
in some situations. This is not a problem
non-ordinary positions needs to be tested
in later generation autonomous cars,
in higher velocities before validating that it
where our target year 2035 is situated. In
could be done. If our project is heading into
2035, infrastructure and cars autonomous
this direction we need to test this.
features are advanced enough to allow

sleeping in the drivers seat.
7.24.3 Dark Horse Prototype - Sleeping

Positions
Since almost everyone would like to sleep
in the car if it would be possible, the team
wanted to build various sleeping setups to
imitate different possibilities to do so. Two
assumptions exists:
1. Driver sleeps in a comfortable position so
that he/she can enjoy complete relaxation.
However, in this situation, it is hard for the
car to wake up the driver in a relatively short
time.
Setup One (Figure 80)

Using the first assumption as a guideline,
the team came up with sleeping position
sideways to direction of movement. Driver
seat turns 90 degrees to the right from the
face forward orientated position and the chair
transforms into a bed. This idea derives from
the fact that beds are positioned like this in
trains to prevent possible motion sickness.
Setup Two (Figure 80)
Under our second assumption, we observed
that if it is possible for a person to fall asleep
Final Documentation
in straight sitting position as long as the

users head is supported so that it does not
fall. Therefore, we made a head holder out
of pillows for the user to see if it is possible
to sleep like this or not.
Setup Three (Figure 80)
Third setup included a device that has a
flat and soft surface to support the head
and the same kind of chest support the
upper body (Figure Sleeping Positions 3).
This setup got its idea from the way people
sleep in airplanes. Since the inclination of
the seats are so minor, sleeping is difficult
in a regular way. In this kind of situation
people are forced to figure out more ways
to sleep comfortably. In airplanes you can
see people sleeping on the table in front of
them leaning on their hands
313
314
Final Documentation
Observations
Setup
The first setup is not much different from
We tried to draw augmented information
the regular sleeping position. Issue with this
on a glass of a protection helmet. Details
setup is to have the required space in cars
we drew on the glass was mainly driving
cabin. Width of the car is not enough for
related, such as speedometer, road and
this kind of setup so it is impossible to sleep
navigation information. (Figure 81)
sideways.
Conclusion / Lessons Learned
In the second test setup the users neck
We quickly immediately realized that this
gets tired really quick since the weight of
technology is almost impossible to imitate
the head is too much to support if the neck
like like this since the eye has to focus on
muscles are operating because of sleeping.
the details drawn on the glass and therefore
This setup is not good for even a quick nap
blurring environment. Therefore this is not
since it causes ache for the user.
useful to imitate augmented reality like this.
In the third sleeping works actually quite

well if the support is just correct for the
user. Spine is more relaxed as well as neck.
Users chest gets a little too much pressure
since the body weight is mostly on chest
and breathing causes chest movement and
increases load on the chest.
Conclusion / Lessons Learned
Replacing the traditional sleeping position in
rational way is not possible since we would
have to compensate with users ergonomics
and we are not willing to do so. Sleeping in
the drivers seat has to be done somewhat
the same way it is done in the seat next to
the driver.
7.24.4 Dark Horse Prototype - AR
Imitation
Because augmented reality technology (AR)
is going be available in future and various
applications are going to be possible with that,
we wanted see if it is possible to imitate it.
Top - Protection helmet split

Bottom - Transparent Foil
Final Documentation
315
7.24.5 Dark Horse Prototype Disappearing Steering Wheel

With this prototype we wanted to test one
of our ideas for transition between, what
we called, manual and autonomous mode.
Idea was to create a mechanism that will
hide the steering wheel, so the user can
utilize the dashboard space for some other
activity, when he/she is not required to steer.
The whole purpose of this prototype was
to visualize in a tangible way what we had
in mind and then, to use that prototype to
inspire conversation or to get some ideas.
Setup
This prototype is one of many quick and
dirty prototypes. We made a cardboard
steering wheel that we attached to a
cardboard table (which imitates the
dashboard - Figure 82) We made two slices
in the table and used the tape as hinges in
order to make the board rotatable.
Observations
When flipped, steering wheel requires a

lot of space below the dashboard.
We could imagine that the cuts in the

table would give the interior very cheap
outlook.
The cuts wouldnt be very convenient,

especially when the table is used for
eating.

Although the prototype was rough we
quickly realized that this solution for making
utilizing the dashboard was not probably
the best solution. This construction gave
very mechanical outlook and caused more
problems than it has solved. We realized
that we need more simple and more
sophisticated solution to this problem.
316
Final Documentation
7.24.6 Dark Horse Prototype -
in some convenient place. Places
Magneto
suggested:
Idea behind Magneto prototype was to test

what users will do with the steering wheel
if it was possible to attach anywhere on the
dashboard. Prototype acted mainly as a
conversation starter.
Setup
Cars roof
On the bench next to drivers seat
In personal bag or purse
What would the user do with detachable

steering wheel?
in leisure mode for cars infotainment
We made a dashboard made out of

cardboard with metal underneath, and
steering wheel of plywood with a magnet in
order to fake the steering wheel mounting.
(Figure 83)Test setup was organised on a
table and mood was set by explaining that
there is two modes of driving: manual mode
where you drive yourself and autonomous
mode, where the car drives itself. During this
we conversed with the tester and tried to
collect valuable feedback from it. (Figure 84)
Discussion Key Notes
Where would the user put the steering
wheel?
User would put the steering wheel
Use the steering wheel as control device

system
Steer the car from outside like a remotely

controllable car
Wheel could have a picture of the car

from bird view like a computer game
What other functions detachable steering

wheel makes possible?
Steering wheel could be used as a

personal tablet
Augmented reality in the wheel, like

personal tablet but with more futuristic
technology
Steering wheel could be a key as well

that could lock the car and prevent its
usage if the steering wheel is not near
Final Documentation
the car
User testing is a fundamental way of getting
feedback and ideas from prototypes.
Prototype does not have to be that
sophisticated to open a conversation and
communicate the idea. Here we learned
that building and testing more is learning
more. Ideas generated by users was listed
317
318
Final Documentation
7.25 Funky Prototypes
chair adjustments before continuing on to
Funky prototype is a term used in ME310 for
to determine if users value this anticipatory
an approximation prototype of the full system

without making a costly commitment to any
one configuration, technology, or geometry.
This prototype is still a low-commitment,
rapidly assembled, concept prototype that
allows for objective evaluation and testing.
a different activity. This prototype was trying

system and if it could be achieved by using
force sensors.
Test Setup
Users were asked to perform the list of
activities taped to the dashboard in any
order and at whatever interval they wanted.
7.25.1 Funky Prototype
The list of activities included driving, trying
Anticipatory Chair
to take a nap, reaching for your backpack in
The team wanted to explore the concept

of facilitating seamless transitions between
activities by having a chair that senses the
changes in your body position and language
& moves to accommodate those changes.
The hope is to create a feeling of more
mobility by removing the need to make
the backseat, watching TV, texting/ reading

a magazine, and talking to your passenger.
(Figure 85) User activities displays the
activities being performed from an actual
Final Documentation
319
320
Final Documentation
User Positions in the Car
user test.
Technical Results
The users were told that the chair would
Three patterns were noticeable in the post-
adjust based on their movements in the
processed force sensor data: sitting up,
chair. There was a button (as shown in
rotating to reach something in the back
Figure 87) to override the actions being
seat, and pushing back against the chair.
made by the chair on the steering wheel,
The following pages explain the data and
which users were instructed to press if the
patterns that emerged. The graphs are
chair ever incorrectly moved or anticipated
arranged, as the force sensors are arranged
their needs. A second switch was on the
on the chair. The top two graphs represent
dashboard so that the user could switch
the top two sensors on the chair, and the two
between windshield mode and watching
right graphs represent the two right sensors
TV. The debrief interview was conducted
placed on the chair when looking straight on
in the test setup, while the chair adjusted
to the chair. (Figure 88)
to better suit this social dynamic. The tests

lasted approximately 20 minutes.
Dashboard Test Setup
Final Documentation
321
Force Sensors
Force Sensors on the Chair
322
Final Documentation
Observations
passengers. In general, users had a positive
Users would pull at the wheel slightly
experience with this anticipatory sensing
when they wanted to move into driving
chair prototype when the chair matched their
position.
intentions. In instances where the chair was
Users would begin to push back against
initially incorrect in its anticipatory action,
the chair if the chair moves in a direction
users still did not want to regain manual
that the user did not want.
control over the chair and felt it made their
Rotating the chair to reach a backpack
transitions easier.
was useful only when the backpack was

out of reach. Otherwise the chair did not
react fast enough to be helpful.
Comfortable angles of recline differed

for each user.
Results
The team discovered that force sensors
seem like a fairly feasible way to identify
simple user intentions like pushing back
on the chair to recline, but a higher density
of sensors in conjunction with additional
types of sensors will be needed in order to
better distinguish between more complex
intentions like the desire to socialize with
Seat Back Data
Sitting up =
0 force read
Seat hitting lower

back as chair raises
into driving position =
INCREASE in force
Force Pattern 1: Sitting Up (from reclined

position to driving position)
These graph (Figure 89) demonstrates
leaning up from the seat back and waiting
for the seat to move into driving position.
When the person sits up, the two upper
sensors read no force since their back is
not touching the chair. Once the seat starts
touching persons back, the two lower
sensors become active.
For the seat cushion, as the person sits up
from a reclined position all their weight is then
shown in the seat cushion force sensors,
which is why one can see an increase in all
four bottom sensors to a steady state value.
Final Documentation
323
Seat Cusion Data
Sitting up =
INCREASE in force
Force Pattern 1
Force Pattern 2: Rotating (to reach for
red mark transition, the data shows a lot
something in the back seat)
of movement shifting. Since there were not
These graph (Figure 90) demonstrate
enough sensors to capture the entire chair,
rotating / leaning towards the back to grab
the variation makes it hard to distinguish
something. At the red transition mark, the
what exactly is going on. Conclusion, need
entire top right side sensors goes to zero
to increase density of sensors among the
while T1 (lower left sensor, the direction
chair to help filter out small movements
the user was leaning) increases. After the
shifts that are unintentional.
Seat Back Data
Top right sensor =

0 force
Lower left sensor

as user rotates
in that direction =
INCREASE in force.
Force Pattern 2
324
Final Documentation
Seat Cusion Data

Rotating to this
side = INCREASE
in force
Rotating to other side = DECREASE in

Force Pattern 3: Pushing Back (to recline)
These graph (Figure 3) demonstrate pushing
back on the seat. The top seat back sensors
(T2 and T4) increase as the user pushes
against the seat back.
The top seat cushion sensors (B2 and B4)
decrease, while the lower seat cushion
sensors (B1 and B3) should slightly increase.
This increase is seen in B3, but because of
the users position in the chair B1 did not
detect this increase.
Seat Back Data

Pushing on seat
back = INCREASE
in force
force on all right sensors
Final Documentation
325
Seat Cusion Data

Pushing on seat
back = DECREASE
in force
Pushing on seat
back = slight
INCREASE in force
Force Pattern 3
7.25.2 Funky Prototype
sceneries, one with a dull road and one with
What should I do
fun-to-drive road. All test situations were
The team wanted to find out gestures and

subconscious movements users performed
within the cabin space to validate whether
people perform unintentional gestures in
unfamiliar situations. This prototype was
suppose to encourage innovative thinking
and offer us new and different ideas for
triggering these two modes. The team
believed that this prototype might uncover
intuitive gestures that people do when they
oriented themselves into a driving or leisure
position.
Test Setup
The concept of the prototype was to design
a dashboard with a steering wheel that
could disappear by flipping the dashboard.
It consisted of a car seat and a dashboard
that had a steering wheel on one side and
an original dashboard on the other side. The
windshield was imitated with two different
filmed with a camera to later observe the

users reactions and body language.(Figure)
The team tested the prototype in several
different locations including Design Factory,
Administration Building, and Motonet (a
local hardware retailer store).
The test started with an explanation of the
situation the user was in. We ended the
explanation by suggesting the user to try
to summon the steering wheel. After the
user got the wheel to appear, we reset the
scenario and asked the user to retract the
steering wheel. In our test there was no right
gesture for summoning the wheel, so we did
not a have a locked gesture for summoning
the steering wheel. The test process is
explained in the picture on the next page.
326
Final Documentation
To give the user a motivation to change
swapped the driving scenario to fun-to-
between autonomous and driving modes, we
drive road and asked the tester to summon
explained to the user that they would have to
the wheel. After the user got control of the
switch modes based on the road conditions.
steering wheel, we changed the boring road
The tester started with autonomous mode on
back to trigger the urge to switch back to
a boring road. Then our facilitator suddenly
autonomous mode as shown in figure on

the right.
Situation
explanation
User doing the

transition to
autonomous mode
User doing the

transition to
manual mode
Debriefing the
test with the
user
Final Documentation
Autonomous mode
Boring
road
Manual mode
Fun
road
327
Autonomous mode
Manual mode
Trigger for
Transition
Fun
road
Manual mode
Autonomous mode
Trigger for
Transition
Boring
road
Test Results
regular buttons were most commonly
We defined inputs that the users gave us
suggested. Most of the hand gesture
as suggestions and categorized them into
suggestions were performed subconsciously.
different groups. There were 11 testers with
This was what we were looking for, but as
suggestions pertaining to both driving and
can be seen from test dat, it was not a
autonomous mode. We have 44 suggestions
common suggestion. Although there was
collected in the table below. To ease the
hand gestures at some point, it was never
comparison between different suggestions,
the first suggestion.
the data is collected into graph that can also

be seen in (Figure 95).
Observations
Testers were afraid to fail, they did not
These were mainly suggestions that users
go wild on ideas as we thought they
gave us knowingly. Voice recognition and
would.
328
Final Documentation
30
Amount in Total
25
1st Gesture
Suggestion Amount
20
2nd Gesture
Suggestion Amount
3rd Gesture
15
Suggestion Amount
4th Gesture
10
Suggestion Amount
5th Gesture
Suggestion Amount
Touch
VoiceG
estureE
ye
Recognition
Brain
Waves
Suggestion Genre
Categories
Touch
Voice
Gesture
Eye
Rec.
Brain
Waves
Suggestions
in Total
Amount in Total
23
11
44
14
19
11
1st Gesture
Suggestion Amount
2nd Gesture
Suggestion Amount
3rd Gesture
Suggestion Amount
4th Gesture
Suggestion Amount
5th Gesture
Suggestion Amount
Final Documentation
If the first suggestion they gave was not
right, they thought quite a while before
Voice recognition and touchable buttons
another try.
Some users lifted their hands to imitate

driving when they realized it is a mock
up of a car interior.
People from Design Factory are a lot

more open to wild ideas and it is easier
to get those from them than people out
of the university area.
It seemed to be really hard for users to

position oneself far in the future.
Users tended to suggest ways that were

already working functions for them. For
example one user was praising voice
control since he had had so good
experiences with his voice controllable
GPS.
Debrief Feedback
One tester would like to have some

indication when the controls are under
his control, other than only steering
wheel appearing.
Users can be thought any kind of

behavior, for example calling to service
lines every time when there is a
problem is a learned habit from work
for example. There is always someone
else responsible for matters that you are
not responsible for.
Test setup was too humiliating, tester

got frustrated when the suggestion he
was sure was right did not work.
329
Voice browsing through menus are

too complicated and slow. Most likely
the voice commands will not be like
that, more like direct commands to do
different actions.
seemed to still be the most intuitive way for

users to transit from autonomous to driving
mode. One possible solution is to have
a combination of commands that map to
different user interaction. Another possible
solution is hand gesture that initiate various
modes. If the user does not have hands
on the steering wheel, the car will clearly
increase autonomous functionality and if
hands are on the steering wheel, the car
will decrease autonomous feeling.We also
realized that especially in car world, the
prototype has to be sophisticated enough to
communicate well with the tester, otherwise
it is really hard to get valuable feedback.
Testing with users outside the university
environment is a whole other matter. The
whole test setup needs to be very well
planed. In Design Factory it does not really
matter, since people are used to test different
kinds of prototypes, but normal citizens who
just want to take care of their daily business
is a different matter. In the future we have
to be really careful that we do not confuse
testers with unclear test setup.
Since the year 2035 is far in the future, it is
difficult for normal people to comprehend.
We have been living in the year 2035 for
4 months and the futuristic world is really
familiar to us. But for random people in a
store it is almost impossible to step into
future just like that. For future testing we
should not include the futuristic aspect too
much, it seemed to just confuse the user.
330
Final Documentation
7.26 Functional System

Prototypes
in the driving mode and is hidden in
prototype setup 1 & 2. The steering wheel
How well can people adapt to this new

chair interaction?
Are gestures an appropriate method to

initiate transition between driving and
autonomous mode?
Functional System prototype is a term used

for ME310 prototype that is still a bit crude,
and obviously assembled from off-the-shelf
parts; however, this time decisions on
technical implementation are done with
increased sophistication.
7.26.1 Functional Sytem Prototype
Anticipatory System
Based on a successful funky prototype, the
team decided to move forward and test the
entire functional system with full integration
of the steering wheel and Chair Sense being
implemented in a closed loop. The concept
being tested is one, which facilitates
effortless transitions between activities by
sensing the intentions of the user through
intuitive body command movements and
intentional mode initiating gestures.
Prototype Description
The funky prototype was developed further
to include closed loop commands for the
chair motions based on inputs from the
force sensors on the chair and gestures
being recognized from the Microsoft Kinect
camera. Another addition to the prototype
is a functional steering mechanism, which
emerges from the dashboard only when
autonomous modes as seen in Figures

96 and 97. Complete functional system
is also hooked up to a driving simulator for
users to get the experience of driving and
transitioning between autonomous and
driving mode. A projector has been used to
mock up the windshield in front of the user,
as they would experience in a real vehicle.
The team also verified that the chair twist/
rotational motions work with a person sitting
in the chair.
Final Documentation
331
user was in autonomous mode.

Test Setup
The user was asked to sit in the system
The gestures that the user asked to perform
and had all of the interactions and gestures
during the test were an calibration pose to
described to them, and a brief introduction
get the Kinect to recognize their body, a
phase occurred so the user could get the
drive pose to move into driving mode, and
hang of all of the gestures and interactions
a retract pose to move into autonomous
with the chair. A driving simulation game
mode. These gestures can be seen on
called Crazy Taxi was being projected
the next page (Figure 98) to demonstrate
onto the screen in front of them as a
the way the Kinect reads these different
windshield and they could take over control
gestures.
of the moving car when the steering wheel

emerged. A team member drove the
videogame car behind the screen while the
332
Calibrated Position
Final Documentation
The user also tried to utilize the chair

sense feature. The team would adjust the
placement of the force sensors in order to
ensure a better data output and to ensure
that the amount of pressure they needed
to exert on the chair to make it incline or
decline was not too much or to sensitive. The
modes they could experience were driving
mode and autonomous mode. Driving mode
entailed having the chair move closer to the
dashboard into an upright driving position
and having the steering wheel emerge.
Drive Position
Autonomous mode entailed having the

chair move away from the dashboard and
recline into a more comfortable and relaxed
seat position, as well as having the steering
wheel retract and be hidden from the users
view and out of their reach. Once the users
were familiar with all of the interactions with
the system, the team let them just transition
through the activities and modes they as
they pleased.
Results
Chair Sense interaction:
Retract Position
For tall or short people it was difficult to

use the interaction.
The force sensors were often not placed

correctly to ensure proper pressure
distribution.
Due to the angle of the Kinect, it was

unable to capture their entire upper
body so the Kinect could not understand
their gestures.
The responsiveness of the sensors was

annoyingly sensitive
The chair was oscillating around the

position the user actually wanted to be
Gestures recognized by the Kinect software 1, 2, & 3
Final Documentation
333
in.
changes. The data collection points also
Pushing back against chair felt very
cannot be fixed to point on the chair, but
natural
rather use reference points from the users
There is a sandwich threshold, which
body. For instance, using pressure data
the user will get stuck in and be unable
where the chair touches just below the top of
to push the chair back.
the users shoulder. That way height will not
If this threshold is met, users immediately
affect the data and chairs responsiveness.
want to turn off these smart features or
Another important discovery was that
get out of the chair.
gestures were triggered randomly just when
People are afraid of being crushed
a person was gesticulating. Therefore, we
by the chair allowing to go past this
do not think gestures should be used as
sandwich threshold
mode initiators. Instead, gestures could be

used as a way of weeding out unintentional
body movements as opposed to intentional
Steering Mechanism Interaction:
body commands.
People wanted to keep both hands on

the steering wheel, while signaling that
In general, people enjoyed the experience
they wanted to go into autonomous
and concept. Specifically, they enjoyed
mode.
moving from driving mode into autonomous
Removing their hands from the wheel
mode as it meant the chair moved to a more
made them feel unsafe and felt awkward
relaxed position and the steering wheel went

away. It made them feel free to not worry
Other Considerations:
about driving anymore and instantly put
Most users wanted a twist motion.
them in a position that was more enticing.
All users described several instances in

which they would want this exist, even
in cars today
Wished there was a way to lock &

unlock chair sense
Users wanted a way to tell the chair to

stop in the position they wanted, in the
event that it incorrectly reacts.
7.26.2 Functional System Prototype Buckle It Out
What is needed for effortless transition?
What kind of problems there are while

changing between autonomous and
driving mode with a seatbelt switch?
The two design teams conducted different

Conclusions
The team concluded that the thresholds we
set for the sensitivity of the chair must be
dynamic and change as the chair position
prototypes for the same purpose: to provide

an easy trigger of autonomous mode and
manual mode, and to test the transition
between mode changes. Both teams
decided to hide the steering wheel when in
334
Final Documentation
autonomous mode, based on the rationale
The team found out that if the steering wheel
that :
was removed, there was 15% more cabin
1) it is less confusing for driver not seeing
space. The concept of this prototype was to
the steering wheel what it is autonomous
design the steering wheel so it folded inside
mode,
the table when the driver freed their hands
2) the self-spinning steering wheel will not
from driving and wanted to use the space
be a distraction when the car is driving
to do something else. The goal was to test
autonomously,
what kind of information drivers needed
3) retracted/folded steering wheel saves
to smoothly transition from one mode to
space for other activities. Two different
another and whether our mode-changing
methods for trigger of changing modes were
trigger (seat belt) was easy to use. The
prototyped.
assumptions were that the key activities in
The first one was the seat belt and the
the car are driving, socializing and working,
second one was gesture control. In addition,
that can be seen in the Figure 99.
the two prototypes provided different

solutions for hiding the steering wheel.
Final Documentation
335
Test Process
to unbuckle the seat belt when the car
The team tested four individuals from
was still moving.
different age categories. Test users were
Seat belt interrupted the activity flow
initially told that they were now sitting in a
when switching to autonomous mode.
self-driving car on their way home, and could
The driver cant let the wheel go while
take control over the car when in driving
unbuckling the seat belt. Here transition
mode. When the car was in autonomous
is not smooth from manual mode to
mode, the steering wheel was folded inside
autonomous mode.
a table and retracted out when in driving
A cooling down time before the mode
mode. A driving simulator was displayed on
change can give driver better preparation
a computer screen so the testers could see
physically and psychologically.
the road condition and take control over the
Trigger for manual auto might be
car. Instead of telling testers how to change
different than trigger for auto manual
mode, the team tried to let the users find the
because manual auto requires
trigger first.
more attention and more complicated

safety system. A more intuitive, easily
Observations
None of the testers realized that seat
accessible trigger is needed.
Autonomous function activation system
belt was the way to summon the
might be different from activation system
steering wheel.
for wheel movement.
They either pulled the wheel, searched

for button, tried voice control or gesture,
Conclusions
etc.
To make the transition between modes as
The team had to tell the testers to fasten
easy as possible for the users, the team has
the seat belt.
to really think through the transition process.
The wheel popping out surprised them.
The process has to be simple and clear for
The team was surprised to see that not
the user. Also the user has to be aware
many people unbuckle the seat belt
all the time of what the car is doing, what
to change mode. Instead, they would
mode the car is in, and how to change the
rather like to push the wheel back.
car to different modes. The same goes with
The team realized that seat belt might
different activities within the cabin.
not be the best way to activate one

mode.
Results
Seat belt as a trigger brought confusion.

Although seatbelt is fastened only when
the driver is driving, driver was confused
336
Final Documentation

7.27.1 Design Specifications For
Chair Electronics Discovery

The chair that was used in the prototype was
from an older Audi car. All the electronics
Anticipatory Chair
to control the seat adjustments were fully
The technical setup of this prototype
functional. The most technical part of this
included the fully functional car seat that was

hacked into to be able to control manually
from a remote area. The chair was fitted
with eight force sensing resistors (FSR)
connected to an Arduino Mega2560 board.
The Arduino was connected to a computer
running Matlab that collected the raw data
from the sensors and the time stamp marks
that notified of the test users transition
between activities. Post processing scripts
were created to construct visual plots of
the sensor data versus time to examine
whether distinct patterns relative to the
users intentional movement positions were
noticeable. If these patterns were obvious,
then having the chair automatically make
the adjustments based on these patterns
could be a possibility.
prototype consisted of making the necessary

adjustments to the chair electronics to
control the adjustments from a remote area.
The first main objective was to figure out the
how the chair worked and where all the motor
and adjustment control lines were routed.
The team re-engineered the seat without
any documentation or electronic schematics
of the seat controller. A pinout diagram of
the seat wiring harness was created to map
out where the signals from each controller
were routed. Figure Connector diagrams
from seat wiring harness 1 & 2 displays
the necessary pin out diagrams of all the
connectors that were under the seat. The
only adjustments that were considered
were the back seat tilt, the bottom seat tilt,
the bottom height adjustment, and the seat
sliding function.
Motor Control
Signals(black connector)
+12VDC
Motor 4 P2
11
N/C
N/C
Connector Diagrams
GND
Motor 3 P2
Motor 2 P2
Motor 3 P1
Motor 1 P2
Motor 2 P1
Motor 1 P1
Motor 4 P2
Final Documentation
337
Motor Sensing
Signals(blue connector)
1
2
6
9
10
12
13
14
15
Red Conn P6
Red Conn P7
ACC Pin4
Motor 4 Encoder
Motor 1/2/3 GND
Motor 2 Encoder
Motor 4 GND
Motor 1 Encoder
Motor 3 Encoder
17
19
22
23
29
Motor 1
(Front bottom tilt)
1
2
Power
Return
Motor 2
(Chair slide)
Encoder
GND
4
3
ACC Pin3
ACC Pin1
ACC Pin5
ACC Pin2
Red Conn P2
1
2
Power
Return
Motor 3
(Back bottom tilt)
1
2
Power
Return
Motor 4
(Chair tilt)
Encoder
GND
4
3
Main incoming / outgoing

(Red connector)
1
6
7
1 - Blue conn P1
2 - Blue conn P28
3 - N/C
4 - +12VDC
5 - GND
Connector Diagrams
Encoder
GND
4
3
3
5
10
6 - +12VDC
7 - Blue conn P2
8 - Seat Belt
9 - N/C
10 - Seat Belt
1
2
Power
Return
Encoder
GND
4
3
Adjustment Controller
Connector
5
1 - Back tilt
2 - Chair slide
3 - GND
4 - Bottom height
5 - Bottom tilt
338
Final Documentation
To Controller Input
The second objective was to figure out

how the switches on the chair worked in
conjunction with the electronic controller.

Figure Switch box controller schematic
S?
SW - SPDT
displays the circuit schematic that was

designed to work with a switch box the
team created. Since the controller circuit
schematics was proprietary and could

not get access to it, the team tapped onto
the seat control adjustment connector
450
700
while hooked up with the controller. It

was observed that no 12VDC power line
was being routed into the connector but a
GND (ground) line was. This immediately
GND
meant that the switches, when closed,
Once the switches were figured out, a
were switching to GND. The next thing
manual switch box was created so that the
that was discovered that the switches were
chair could be controlled from a remote
momentary single-pole double throw and the
area rather than from the switch controls
resistances on each of the throws could be
on the chair. Figures Switch control box 1
measured. One resistance was measured
& 2 display the manual switch box that the
as ~700ohms while the other resistance
team designed.
was measured as ~450ohms. Since only

five signal lines were routed into the seat
adjustment connector and one was GND,
the other four signal lines corresponded
to the four different switches that could be
activated. Each switch has two options,
either forward or backwards direction. This
led the team to uncover that the controller
used an analog-to-digital converter onboard
to be able to sense whether the voltage from
the switch side corresponded to the 450ohm
or the 700ohm side of the switch, resulting
in the motor to spin forward or backward.
Final Documentation
339
Force Sensing Resistors

The chair was outfitted with eight FSRs to
collect the users weight distribution data
across the chair. Figure 103 displays the
chair with the sensors and the nomenclature
used to distinguish each sensor for data
processing. The FSR were connected with
connectors to make it easier to debug/
repair each individual sensor. A proto-shield
that connected to the top of the Arduino
Mega2560 was used to breadboard the
FSR circuits. Figure FSR proto shield the
complete proto-shield and Figure Schematic
for one FSR circuit displays the circuit
schematic of how each individual sensor
is connected.
Switch control box 2
Force Sensors
Position of the Force Sensors
340
: FSR Proto
Final Documentation
Schematic for one FSR circuit
+5VDC
R6
FSR
3
R7
3.3K
GND
The FSRs are essentially potentiometers

(variable resistors). The amount of force
applied to the sensor pad is inversely
proportional the amount of resistance across
the terminals of the FSR. Within the circuit
that was used, the more force applied,
a larger voltage drop is seen where the
connection to the analog input pin is tapped.
The onboard ADC was used to convert this
voltage of 0-5V range to a 10-bit digital
value. A higher digital value corresponds to
more force applied. The Arduino collects the
data from a each sensor approximately 40
times per second and sends the raw data via
serial communication to Matlab for further
post processing.
During initial user testing, it was observed that
the density of FSRs that were implemented
were not sufficient in determining the weight
distribution of the user over the entire chair.
The FSRs only detect how much pressure

is applied to the pad and not where the
pressure is being applied to on the pad.
For FSRs to work, higher density of sensors
needs to be placed in localized areas to
obtain a better weight distribution.
MATLAB Processing
MATLAB was used, via serial communication
with the Arduino, to collect the raw data
from the sensors and the timestamp
that marked the data to notify the users
transition between activities. MATLAB
saved the FSR data and time stamp in text
files. Post processing scripts (Appendices)
took the information within the text files
and constructed visual plots of the data,
separated by top and bottom. The team
analyzed these plots and the user test
videos together to verify that the users
Final Documentation
341
visual data represented the patterns in the
plot. Figure Plots of FSR data from a user
Anticipatory System
test displays a users test data plotted for

analysis. The plots were constructed to the
same nomenclature outline to make it more
intuitive.
This prototype system consists of 5

subsystems that all work in conjunction
to provide the user with an interactive
responsive chair experience as seen in
Figure Subsystem component diagram.
These subsystems include the Microsoft
Kinect (IR Camera), Force Sensors and
Board, Chair Rotation Mechanism, Steering
Wheel Mechanism, and Motor Relay Driver
with Motor Position Detection. The Microsoft
Kinect used and coded to detect specific
poses or gestures that would indicate
whether the user wanted to be in driving
mode or autonomous mode. The force
sensors and board were used to collect
Plots of User Test Data
the data used to determine chair tilt body

intentions to recline or incline the seat back.
The chair rotation mechanism allowed for the
chair to rotation approximately 20 degrees
to help users sitting more comfortably
while talking to a friend or reach something
from the back seat. The steering wheel
mechanism allowed the steering wheel to
be hidden when not in use and to emerge
when the user wants to drive. And lastly,
the motor relay driver and motor position
detection would enable motors in order for
the chair to reactive to these user (gesture
and intentional body commands) and force
sensor inputs.
342
Final Documentation
IR Camera (Kinect)
Force Sensor Board
Arduino
Chair Rotation Mechanism
Steering Wheel Mechanism
Motor Relay Driver
Motor Position Sensing
Reley Driver Circuit Schematics
Relay Driver Board
motors on the seat, one motor to control
The team designed a custom motor driving
the steering mechanism, and one to control
and position-sensing controller to allow

the closed loop functionality. The OEM
controller that came with the chair was
not used because modifications to the
microcontroller could not be accessed.
The team re-engineered the controller by
visual inspections of the unit and determined
how exactly the controller knew where the
position of the chairs motors were at all the
time.
To be able to gain control over the motors,
H-Bridge circuits were created out of
DPDT switches to allow the motors to spin
backwards and forwards, as shown in Figure
108. SPDT switches were used to connect
+12VDC or GND to supply power or lock
the motor. The circuit schematic displays
only one relay setup. Since there are four
the rotation mechanism, six individual relay

circuits were designed in parallel on a
protoboard, as shown in Figure Relay driver
circuit schematic and board 1 & 2. SPDT
switches were used to connect +12VDC or
GND to supply power or lock the motor.
Load current measurements were obtained
Final Documentation
343
+12VDC
Enable power
D1
1N5817
GND
K1
ORWH-SH-105H-3F
(Relay SPDT)
D3
Motor control
GND
D3
1N5817
K2
RTE24005F
(Relay DPDT)
GND
M
Motor
prior to selecting parts to build the relay
the forward and backward positions back
board. Load currents while the motors
electromagnetic field spikes are generated.
were running were approximately 3-4A
These spikes can potentially damage parts
with an average weight of 120lbs. As the
and disrupt/create noise on sensitive logic
motors reached and stalled at their running
signals in the nearby area.
limits, the stall current was approximately

9A. The DPDT and SPDT switches were
During initial testing of the entire prototype,
chosen with current ratings of 10A and with
the team discovered that to create a
a switching coil voltage of 5VDC. This coil
more refined experience for the user, the
voltage was pertinent since the board was
transitions between chair positions must be
designed so the microcontroller logic levels
smooth and non-intrusive. The on/off control
would be able to control the switches. A 10A
signals to the motor did not provide the
current rating seemed reasonable since the
experience that the team ultimately desired
team knew that software would prohibit the
so future designs of the relay board will
motors to run until they stalled out. 1N5817
replace the SPDT switches, that supplied
Schottky diodes were applied across the coil
power to the motor, with Darlington power
control inputs and between +12VDC and
transistors and appropriate heat sinks to
GND. Since motors and coils have inductive
dissipate the wasted energy. The relays
properties, as the motor switches between
had a switching time of approximately 10ms.
344
Final Documentation
Although PWMs could be used to control
signal line was tapped to the encoder sensor.
the motors, the motors did not run properly
With the use of a potentiometer in series with
with such a low PWM frequency since the
+5VDC that hooked to the encoder line, the
motors required a much higher frequency to
optimal resistance value (~440 ohms) was
run without sharp transition steps.
found that allowed a low frequency squarewave pulse to be transmitted on the line.
Motor Position Sensing Board
The square-wave pulse had a DC offset of
The next step after gaining control over the
~2.5V and a range from 2.5V to 5V. These
motors with the relays was to determine how
parameter ranges were not sufficient to use
the original seat controller knew the positions
with the Arduino logic levels so a comparator
of each motor. The team knew there was
setup was used with a threshold detection
some kind of encoder sensor on the motor
of 3V, as shown in Figure Motor position
but did not know how it was possible since
sensing circuit schematic and board 1 &
encoder sensors usually consist of 3 signal
2. This allowed the output to step between
lines, VCC, GND, and the sensing line.
0V and 5V for an input range of 2.5V and
The motor pinouts only had two lines for
5V. The circuit schematic only displays the
the encoder, some signal line and GND.
setup for one motor, but the board houses
After several different attempts, the team
four separate motor sensing circuits for the
discovered that power was being supplied
chair motors as shown in the figure below.
through the encoder signal line while the
Motor Position Sensing Board 1
Final Documentation
345
R2
1K
GND
+5VDC
R2
1K
R4
R5
+5VDC
3
6
1K
470
+5VDC
U1A
LP339
R3
Res1
OUTPUT
1K
12
GND
GND
Figure 111: Motor Position Sensing Board 2
Force Sensing Resistor
placed in areas that would be beneficial to
The same square FSR and circuit schematic
obtain weight distribution data from. The
were used as in the funky prototype. Four
sensor layout can be seen in Figure FSR
additional FSR, a total of 12 FSR, were
layout on the chair 1 & 2.
T1
T2
T3
T5
T4
T6
B1
B2
B3
B5
B4
B6
FSR Layout on the chair
346
Final Documentation
FSR Board
Instead of the FSR circuits attached to an

Arduino protoshield, a dedicated board was
built. The previous protoshield required eight
analog inputs to be used on the Arduino.
The best way to utilize I/O ports on the
microcontroller was to design a board that
incorporated a 16-to-1 analog multiplexer
(with digital control lines) to select between
each of the 12 FSRs. This multiplexer
only required one analog input rather than
directly connecting each FSR to an analog
input. Figure FSR circuit schematic and
board, displays the newly re-designed FSR
board.
CD4067BE
OUT
T2
T1
B6
B5
B4
B3
B2
B1
CONTROL_A
CONTROL_B
GND
OUT
IN7
IN6
IN5
IN4
IN3
IN2
IN1
IN0
A
B
VSS
VDD
IN8
IN9
IN10
IN11
IN12
IN13
IN14
IN15
INHIBIT
C
D
+5VDC
T3
T4
T5
T6
GND
CONTROL_C
CONTROL_D
TI - CD4067BE
FSR circuit schematics
Kinetic Camera Software
to distinguish between unintentional and
The Kinect camera from Microsoft was
intentional movements. One way to help
used in conjunction with the FSRs in order
filter out the noise and small movement
to better detect the body positions and
adjustments was to use another sensor (the
movements of the user. During the funky
Kinect camera). The Kinect has been a very
prototype user testing, the team determined
popular device to develop gesture based
that the FSR data was not accurate enough
recognition applications for a variety of
Final Documentation
347
purposes. The team used the Kinect to get
During initial testing, it was observed that the
specific body positions and gestures related
Kinect did not track the users body position
to enabling driving and retracting mode, but
accurately and efficiently as if the Kinect
also to detect whether the user wanted to
was mounted perpendicular to the user.
grab something from the backseat. Once the

data is processed on the laptop, the state of
Integrated Development Environment
which the users body position is transferred
Setup
via serial communication to the Arduino that
The Kinect device and software was
controls the seat.
used on Mac OSX. Since it is a Microsoft/

Windows based product, third party open
Kinetic Position Setup For Testing
source develop environments were used
The Kinect is setup as shown in Figure
to create the code. OpenNI/NITE was the
Kinect setup FOV. With the device above
open source API library that was used to
the person and tilted downward, the depth
gain access to most of the Kinect device
information for the joint positions must be
functionalities. Processing, a graphical
translated/rotated around the x-axis by
interface that was developed in the same
the amount the Kinect is tilted, which is
manner as the Arduino IDE, was used with
performed in software. This will emulate
the SimpleOpenNI library to develop the
the coordinate system to be similar as if the
data processing code passed to the Arduino
Kinect was mounted directly in front of the
controller. The SimpleOpenNI library for
user within the FOV (Field of View).
Processing is a simple wrapper library that

maps simple call functions to the more
complex API library functions within OpenNI.
Software Description and Functionalities
The software was designed to detect
Kin
ect
specific body positions and gestures. The
FO
team incorporated detection for transitioning

into driving mode (summoning the wheel),
transitioning
into
autonomous
mode,
(retracting the wheel), and turning/reaching

back toward the backseat. The software was
designed around the code from Making
Dashboard
Things See by Greg Borenstein.

The software is split up into two different
files, the functional_system_prototype.
pde (the main file) and the skeletonposer.
Kinect setup FOV
348
pde (checking function file). The main

file initializes all the necessary data
Final Documentation
Calibrated Position
processing events, adds the different body

position rules for detection, and does the
high level checking of each state. The
skeletonposer file implements the rule
functions that are called within the main
function and checks each rule to give a
binary result of being either PASSED or
FAILED. The rules are based on joint-tojoint coordinate comparison. For example, if
the y-coordinate of the right hand is greater
than the y-coordinate of the right elbow and
right shoulder, it can be concluded that the
user is raising their right. Figure 63 hand
Drive Position
shows two of the body positions that were

implemented in software to be detected. The
code checks the rules against the updated
coordinate comparisons at every iteration
of the code loop. Once the skeletonposer
file determines if the rules have PASSED
or FAILED, the main file will send the
Arduino that controls the seat adjustments
a character via serial communication to
represent the state of the user. The states
are decoded as R for retract, D for drive,
and Q for default state in which the user is
in a neutral position.
Retract Position
Once the software is running, the user

uses the calibration body position Psi,
which most resembles the dont shoot
me position as shown in Figure Gestures
recognized by the Kinect software .The
Kinect always has to be calibrated to the
users body for skeleton tracking to be
accurate and enabled.
Gestures Recognized by Kinect Software
Final Documentation
349
Rotational Mechanism
The rotation mechanism allowed for users
to reach the backseat for an item or for
parents to reach back and unbuckle their
childs car seat. Placing the chair base on
a ball bearing turntable, which was also
attached to a motor base, compromised
the rotation mechanism as seen in Figure
Rotational mechanism subcomponent. This
motor base was where we also mounted our
motor and gear. Attached to the chair base
was a large gear, which interacts with the
gear mounted to the motor. This system
only allowed for a 30-degree turn, and the
gears .5 thick laser-cut acrylic pieces as
seen in Figure Laser cut gears for rotational
mechanism. Gear profile information and
motor specifications can be found in Section
Laser Cut Gears
9.10.4 Appendices.
Rotational mechanism subcomponent
Motor
Seat Base
Motor Gear
Seat Gear
350
Final Documentation
Steering Mechanism
The team designed a retractable steering
wheel using a crank slider mechanism. The
main concept is that the steering wheel lies
flush against the dashboard when not in
driving mode. This default configuration has
been shown in Figure Steering wheel tucked
away (disappeared) when in autonomous
mode. There is also a flap, which the users
can use as a work desk if they want to use
that space in autonomous mode to perform
some activity.
As soon as the controller gets a driving
another addition to the functional prototype
mode command it initiates the chair motion
compared to the Funky prototype. The whole
first. Once the chair is in the driving position
steering assembly lies on a slider the base
the steering wheel motor turns for a fixed
of which is connected to the platform on
time interval (since the motor used for this
which the entire setup has been mounted.
prototype does not have encoders on it) and
The motor has been mounted in a laser cut
the steering wheel pops out for the users to
acrylic structure, which is attached vertically
drive. In this prototype, the steering wheel
to a support that goes over the mechanism.
has been connected to the actual driving
The two links and their fasteners have been
simulator steering wheel so that users
chosen so that there is no interference
can actually drive on the simulator, which
during motion. The crank slider mechanism
is being projected in front of them. Giving
has been shown in Figure Crank slider
users this simulated driving experience is
mechanism for steering wheel.
Crank slider mechanism for steering wheel.
Final Documentation
351
Arduino Code Description
the backside of the chair. It is a 10-bit ADC
The motion control module has been
so the force sensor values are going from 0
implemented using the Move Motor
to 1024. It was found through initial testing
functions. It takes as arguments the
that the normal value of this force sensor
distance to be moved in terms of encoder
reading on the back of the chair is around
ticks and the direction of motion. There
450. So the upper and lower thresholds for
is a moveComplete variable on both the
the algorithm were then set as 150 and 700.
motors which is set to 0 when the motors

are moving. This functionality can be used
The software has been implemented in
to ignore motion commands when the chair
the form of a state machine, which keeps
is getting adjusted for a certain mode.
checking for gesture recognition commands

being sent from the Kinect. These commands
There is a motor stall check Boolean function
are in the form of characters received on the
that detects that a motion command is being
Serial port of the Arduino. As soon as a new
applied and the encoder line is not ticking.
character is detected the code feeds through
When a motor stall is detected the controller
the corresponding motion commands to the
immediately shuts down that motor.
two motors. The force sensor smart algorithm

has an override over this functionality except
The Arduino code starts with a calibration
when in drive mode. If at any point the force
function that maxes out the two motors in
sensor reading crosses the upper threshold
one limit position. This is detected using the
then the algorithm initiates backward motion
motor stall function. Once both the motors
of the chair tilt and vice versa for tilting
are maxed out the code then updates the
forward. If the algorithm detects a low on
EEPROM values of the motor position to
the force sensors while moving back then it
zero. The EEPROM is used to constantly
stops the motion in that direction and goes
store the current motor position, so that
to default mode. Figure Smart chair mode
the next time the code starts it knows what
thresholds displays the thresholds for the
position the seat is in currently.
chair positions.
All the force sensors on the chair are

connected to the ADC pins on the Arduino.
The motor encoder lines are connected to
the External Interrupt pins corresponding to
INT4 and INT5. Only two motors are being
currently used for this prototype the front/
back and the chair tilt motors. The Arduino
code is running in a loop and detecting the
force sensor values from the top portion of
352
Final Documentation
R
Steering
Wheel
D
Smart
Chair
R
Smart chair mode thresholds

Buckle it out
System Setup
The prototype is meant to test with the
users for the experience of switching
between autonomous mode and manualdriving mode. The system consists of two
sub-systems that make the steering wheel
to appear. The sub-systems are mounted
to a frame. The sub-systems are pushing
out mechanism and steering mechanism.
Fastening a modified seat belt triggers all
this activity. The casing of the servo was
designed with detail to fit the ends of the
custom made steering instrument and the
rotating servo. After that the electronics and
the software are explained with care. The
final prototype design is shown in Figure
122. Figure 123 shows user tests with the
final prototype.
D - Driving Mode
Q - Default
R - Retract Mode
Final Documentation
Final Prototype Design (top and middle)
353
354
The Frame
The base of the prototype was a chair from
a Fiat Doblo that was bolted to a wooden
frame. The measurements of the mock-up
cabin were taken from the Audi A8. Figure
124 displays a member of the group setting
the layout.
The frame was built already for an earlier
Funky prototype, but the system to control
the steering wheel was incorporated. The
team had two ideas how to make the steering
to appear as shown in Figure 125. One idea
was to have a cylinder-rolling dashboard
that had the steering wheel appear from the
bottom of the dashboard between the legs
of the driver. The concept was initially tested
but was too large to fit in a normal sized car.
The team took measurements from a BMW
Final Documentation
500 taxi and validated that in this concept

steering wheel would hit drivers legs. The
cabin space approximations can be seen in
Figure 126. The other idea consisted of a
joint-like bar that brought the steering wheel
inside the dashboard and opened up like
a flower for the driver (Figure 125 middle
right). In the end, the team chose a solution
related from the original ideas and combined
the two ideas to be appearing steering
wheel that came out of the dashboard, but
retracted vertically in the very end of the
cycle of switching modes.
Final Documentation
355
48cm
24cm
64cm
38cm
77cm
50cm
356
Final Documentation
Steering Mechanism
Gaming Steering Wheel
The steering mechanism was built on
its ready to be driven. Inside the casing
ordinary gaming steering wheel where the
there is a servo and the wiring for it. The
original wheel was cut and the wires were
wheel is a square shaped bent pipe that is
lengthened. (Figure 127) The new steering
mounted to a servo with a friction coupling.
wheel was designed to appear from the
The servo casing was made out of
dashboard. After a flipping movement, the
prototyping plastic by milling. (Figures 128
wheel appeared in front of the user, showing
and 129)
Servo Casing
Final Documentation
Steering Wheel Mechanism
357
358
Final Documentation
Pushing Out Mechanism
by motor controller that is able to handle
The appearing wheel was decided to
24V and more than 10 Amps. The motor
build with a normal gaming steering wheel
controller was controlled by Teensy 2.0++
mounted to a table that moved linearly
with PWM signal. The motor was operated
for a determinate amount. The prototype
blindly, so no feedback was given back to
consisted of a bar that was connected to a
the microcontroller and the calibration was
rolling plate on a table while the other end
made with time sequence that how long time
was connected to a separate small piece of
of operating the motor to a certain direction
table on guides from an Ikea drawer. The
would fulfill the need of 22 cm of linear
construction was made with a 24V DC motor
movement in the end of the bar. (Figure 131)
connected to the bar that transfered the

rotational movement to a linear movement.
The power was supplied from a controllable
power supply. The motor was controlled
Pushing Out Mechanism
Final Documentation
359
Seat belt switch
Teensy 2.0++ Code Description
The Seat belt switch was build of a standard
Code used for controlling the motor can
seat belt lock and embedded touch-sensitive
be found from the appendix. The code
switch. A hole was drilled on the back of the
has a main loop that is a modified state
casing and the switch mounted with hot glue
machine that triggers from the seatbelt
to be pressed by the existing mechanism of
switch. The button has a bouncer control
the lock. The switch was then connected to
that it eliminates all the false alarms from
the Teensy 2.0++ pins with pull-up resistors
the interference of other electronics. The
and polled in the program code. (Figure 132)
calibration is measured by hand and all
Electronics
The electronics on breadboard contains
Teensy 2.0++, a push button for debug,
wiring for the seat belt switch, motor
controller(pin 16) and servo(pin 15) wiring.
the values are timed with the mechanical

feedback of the system.
360
Final Documentation
Final Documentation
361
Thank you for your time!
362
Final Documentation
ME310 2013 Final Documentation

Audi Team

AudiEvolve Documentation 2013

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

AudiEvolve Documentation 2013

Enviado por

Direitos autorais:

Formatos disponíveis

EV

Transforming the journey into the destination

ME310 2013 Final Documentation

1.1 Executive summary

Your autonomous car

activities in the cabin space. The car seat

interactions with the system Mode Initiators

and the steering mechanism are important

(MI) are intentional actions within reach like

components in the cabin space, which will

pulling the steering wheel to go into manual

affect the freedom of motion and ability to

driving mode and pushing on the steering

perform various activities in a comfortable

wheel to make it retract. The user can also

way. The team envisions that the system will

use Intentional Body Commands (IBC) like

consist of a smart chair and a retractable

leaning back to tilt the seat. The chair can

steering wheel. The main motivating factors

also give Adaptive Chair Reactions (ACR) to

for the retractable steering design were

movements of the users like rotating slightly

increasing mobility within the cabin space

when they reach for something at the back

and relieving the users of the responsibility

of monitoring autonomous driving when in

This new cabin space experience allows for

autonomous mode. While the steering wheel

the driver to rotate 180 degrees to be able

is retracted and locked for autonomous

to socialize with other passengers. Having

mode, the tablet is enabled so that the user

only one seat in the front also increases

can interact with the windshield. When the

the mobility within the cabin, while also

steering wheel is not retracted it is free to

making the drivers seat very desirable.

rotate so that the driver can have manual

This configuration allows the driver to

control of the car. There are three different

better share the driving experience with

Figure 1.1.1: Final Prototype

the passengers since it increases their view

and artificial intelligence. With so much

of the road as well. Our final prototype will

technological development in these major

give the experience of facilitated activities

fields there was bound to be a point when

through increasing mobility and effortlessly

these ideas crossover and start influencing

transitioning between activities for the driver.

automotive design. As technology develops

Going forward, the team will incorporate

further, it will lead to a complete paradigm

many of the findings uncovered in all of the

shift in the way cars and automotive

prototypes, research, and needfinding into

technology in general is perceived amongst

a functioning system. The development

the general population. Concepts which

described herein leads to a vision combining

were just science fiction a few years back,

many of the concepts explored during the